Political Parties and Welfare Associations

Home , AS Party, Bible Belt (Norway), Commercial (First), Democrats in Norway, New Party (Turkey, 1993), Political colour

Department of Sociology Umeå University

Political parties and welfare associations

Ingrid Grosse

Doctoral theses at the Department of Sociology Umeå University No 50 2007

Department of Sociology Umeå University Thesis 2007

Printed by Print & Media December 2007 Cover design: Gabriella Dekombis

ISSN 1104-2508 ISBN 978-91-7264-478-6

Grosse, Ingrid. Political parties and welfare associations. Doctoral Dissertation in Sociology at the Faculty of Social Sciences, Umeå University, 2007.

ISBN 978-91-7264-478-6 ISSN 1104-2508

ABSTRACT

Scandinavian countries are usually assumed to be less disposed than other countries to involve associations as welfare producers. They are assumed to be so disinclined due to their strong statutory welfare involvement, which “crowds-out” associational welfare production; their ethnic, cultural and religious homogeneity, which leads to a lack of minority interests in associational welfare production; and to their strong working-class organisations, which are supposed to prefer statutory welfare solutions. These assumptions are questioned here, because they cannot account for salient associational welfare production in the welfare areas of housing and child-care in two Scandinavian countries, Sweden and Norway. In order to approach an explanation for the phenomena of associational welfare production in Sweden and Norway, some refinements of current theories are suggested. First, it is argued that welfare associations usually depend on statutory support in order to produce welfare on a salient level. Second, it is supposed that any form of particularistic interest in welfare production, not only ethnic, cultural or religious minority interests, can lead to associational welfare. With respect to these assumptions, this thesis supposes that political parties are organisations that, on one hand, influence statutory decisions regarding associational welfare production, and, on the other hand, pursue particularistic interests in associational welfare production. It is hypothesised that political parties attempt to mould statutory decisions on associational welfare provision in accordance with the interests of associations with “congruent constituencies.” The aim of this thesis is to examine whether political party preferences for certain welfare associations might help to explain variations in statutory support for associational welfare provision. Two questions are raised: First, do parties differ in their attempts to influence statutory subventions and regulations of associational welfare provision, resulting in more or less favourable conditions for associational welfare? Second, do parties systematically differ in their policies with regard to more or less “congruent” associations? In order to investigate these questions, a comparison is made between political parties’ attempts to influence statutory regulation and subvention of Norwegian and Swedish associations active in the areas of day-care and housing. For this purpose, information is drawn from public documents and official statistics in order to identify more or less favourable policies and related partisan policies. In addition, supportive parties and favoured associations are compared with regard to their “constituencies.” The findings partly support the hypothesis. Although political parties partly pursued consensually association-friendly policies, they often varied their support for welfare associations, whereby both right-wing and left-wing parties partly advocated and partly rejected association-friendly policies in a conflicting way, resulting in varied degrees of statutory support. Furthermore, supportive parties shared “congruent constituencies” with those associations supported by respective parties’ policies. These findings indicate that partisan policies indeed make a difference for associational welfare production, whereby parties of any political colour can support associational solutions. Furthermore, partisan policies vary according to the involved associations’ more or less “congruent constituencies,” which can pursue welfare production out of various particularistic interests, be they religious-cultural or socio-economic in nature.

Key words: Sweden, Norway, Scandinavia, Third sector, Voluntary organisations, Non-profit organisations, Cooperatives, Welfare policies, Childcare, Housing

Now it is pleasant to hunt something that you want very much over a long period of time, being outwitted, outmanoeuvred, and failing at the end of each day, but having the hunt and knowing every time you are out there, sooner or later, your luck will change and that you will get the chance that you are seeking. But coming in at noon, up since two hours before daylight, with only three days left, there, at the table under the dining tent fly, talking away, was Kandinsky of the Tyroler pants. “Hello. Hello”, he said. “No success? Nothing doing? Where is the kudu?” I had forgotten all about him.

- Ernest Hemingway, Green Hills of Africa

Acknowledgements

I want to thank all those who have helped and encouraged me to prepare this thesis. In particular, I am grateful for the support I received from the Department of Sociology at Umeå University. The department welcomed me warmly and helped me generously with all the practical support a doctorand could wish for. Kind financial support was provided by several organisations: The European Commission’s Training and Mobility Program (TMR) for Young Researchers on “Family and the Welfare State” financed a research stay at the University of Gothenburg in 2000, and the former Nordic Academy for Advanced Studies (NorFA) financed a research stay in Oslo in 2002. The Swedish Research Council (VR) supported my PhD project with a grant from 2002 to 2005. The Department of Sociology at Umeå University provided additional financial support. Furthermore, I would like to express my gratitude to all those organisations and their employees, who offered advice and were friendly and patient in helping me find the information I was searching for: the Norwegian Social Science Data Services (NSD), the Information Centre of the Swedish Riksdag, the Norwegian Ministries' Information Corner and Press Centre, the Norwegian Storting's Information Service, the Norwegian Federation of Co-operative Housing Associations (NBBL), the Riksdag Library, the National Archives of Sweden, Hyresgästernas Sparkasse och Byggnadsförening (HSB), Riksbyggen, Norges Kvinne- og Familieforbund, Swedish Social Science Data Service (SSD) and Den Norske Stats Husbank. Of course, I am very indebted to my supervisors, Jonas Edlund at the University of Umeå, Rafael Lindqvist at the University of Gothenburg and Apostolis Papakostas at Södertörn University College. They offered helpful comments on various drafts of my thesis and encouraged me throughout the writing process. Likewise, I would like to thank Dietmar Rauch, Lars Gulbrandsen, Erling Annaniassen, Håkan Lorentzen, Gunnar Olofsson, Jan- Erik Lane, Ulla Björnberg, Einar Överbye, Calle Hult and Lena Karlsson for information, comments and discussions.

Einen großen Dank an meine Eltern, die immer der entscheidende Rückhalt in meinem Leben gewesen sind.

Ingrid Grosse Umeå, December 2007

Contents

Abstract Acknowledgements

I. INTRODUCTION 3 1. Welfare associations: Definition and assumptions 4 2. Earlier research on the preconditions of associational welfare provision 9 3. Statutory and associational welfare activity 20 4. Political parties 40

II. DAY-CARE IN SWEDEN AND NORWAY 49 1. Introduction: Day-care 49 2. Associational limitations and statutory intervention 55 3. Statutory intervention and associational provision 58 • 4. Associations in day-care 70 5. Partisan policies 85 6. Parties and associations: Christian People’s Party, Housewives’ Federation and denominational associations 98 7. Day-care: Summary of main findings 110

III. HOUSING IN SWEDEN AND NORWAY 114 1. Introduction 114 2. Associational limitations and statutory intervention 122 3. Statutory intervention 126 4. The involved associations: cooperatives of differing character 141 5. Partisan policies 151 6. Parties and associations 165 7. Housing: Summary of main findings 175

IV. DISCUSSION AND CONCLUSION 179 1. Introduction 179 2. Partisan policies concerning welfare associations 179 3. Partisan interests in welfare associations 188 4. Outlook: Changing parties and associations 197

APPENDICES 206 Appendix A 207 Appendix B 228 Appendix C 233

NOTES 238

REFERENCES 260

2 I. Introduction

In Western Europe virtually all societies have institutionalised welfare provision for their populations on historically high levels, but they have done so in quite different ways. Thus, contemporary European welfare states differ significantly in, for instance, the overall volume of welfare expenditure (see e.g., Wilensky 1975) and the existence and construction of welfare entitlements (see e.g., Esping-Andersen 1990). Another major dimension of welfare state variation, which, as yet, has received less attention, is the composition of welfare provision among various provider types (see e.g., Alber 1995). In a number of welfare states the provision of welfare benefits is mostly organised by the state or the public sector, whereas, in other welfare states non-public for- profit provision is predominant or non-public associational welfare provision plays a major role. In other words, there are not only discernable differences in welfare expenditure and legal entitlement to welfare provision, but also perceptible differences in operational welfare organisation. These differences can be seen as salient aspects • of variation between diverse paths of welfare state development. In other words, not only differences in welfare expenditure and legal entitlement to welfare provision but also differences in operational welfare organisation can be seen as a salient aspect of variation between different paths of welfare state development. The question that this study will focus on concerns the relative importance of welfare organisation by associations, with a particular emphasis on the situation in the two Scandinavian countries Sweden and Norway. A closer examination of associational welfare provision in these countries in the areas of childcare and housing will serve to answer the question: Under which circumstances (and why) can significant associational participation in welfare provision be established in certain countries and certain welfare areas? Sweden and Norway are interesting test cases when questioning traditional accounts of associational welfare provision. Most of these accounts would hypothesise that, because of their strong statutory welfare involvement, their strong working-class organisations and their ethnic and religious homogeneity, associations in the Scandinavian countries are very unlikely to be involved in welfare provision at all. Yet, in the area of childcare and housing, we can find higher levels of associational welfare provision in Sweden and Norway than in other 3 Western European countries – countries that would be assumed to be more predisposed towards associational welfare. In this thesis I will show that the traditional accounts’ failure to explain this phenomenon is mostly due to an oversimplified understanding of the role of partisan politics behind statutory welfare policies in relation to welfare associations. The notion proposed here is, that political parties of various political colours can be interested in associational welfare production. I will further hypothesise that political parties mould statutory policies concerning associations in accordance with their particularistic interests towards welfare associations, which have “congruent constituencies.” In this introduction I will first define the main concept, “associational welfare”. Thereafter, I will present and discuss the traditional accounts of associational welfare research as a background to the development of my own approach. This approach will be elaborated upon in the fourth and fifth section of the introduction, where the relationships between state, associations and parties will be discussed. The introduction will be followed by the first case study, which examines associational childcare provision in Sweden and Norway. Thereafter, the second case study, which focuses on associational housing provision, will be presented. In the final chapter, the main results will be summarised and a conclusion drawn.

1. Welfare associations: Definition and assumptions In the following section, I will, firstly, develop a definition of the concept welfare associations, as these are the main objects of the study, and, secondly, I will present a number of general assumptions concerning the reasons why associational founders may be motivated to engage in welfare provision.

Definition Welfare associations are defined as formal organisations, that carry out (1) welfare production (2) with profit limitations (3), and are controlled privately (i.e., non publicly) (4).1 (1) By formalising their purpose and organisational structure via written, legally binding statutes, formal organisations distinguish themselves from more informal or ad hoc support networks. In the case of welfare associations, the designation formal organisation 4 distinguishes them from neighbourhoods, kinships, families, friendships, et cetera.2 Formal organisations usually define in their statutes various decision-making procedures and criteria, for example, procedures for how board representatives are appointed, decision requirements of two-thirds majorities or veto rights. Furthermore, decision-making criteria are often defined for matters such as who is eligible to be appointed to the board. Membership-based organisations often apply democratic rules for appointing board representatives, whereas individual-owner organisations may appoint board members in a top- down fashion. Formal organisations usually define various positions according to functional rights and duties, such as administrative and operational positions and professional or non-professional positions. This usually implies some form of hierarchy between positions, i.e., a differentiation according to the influence of various positions in decision-making processes. (2) By formalising their organisational purposes, organisations bind and delimit themselves with respect to certain activities. In this study, • only organisations with the purpose of welfare production in a narrower sense are considered. Organisations such as those concerned primarily with sports or cultural activities are not considered. Therefore, the criterion of welfare purpose distinguishes welfare associations from sports clubs, cultural associations, art foundations, and so on. (3) Furthermore, only organisations that place restrictions on the use of profits resulting from their welfare operations are considered. By stating profit limitations, organisations usually attempt to assure that profits are only used for re-investment into welfare production and not for rewarding owners or stakeholders. The criterion of profit- limitation serves to distinguish welfare associations from private business organisations that act as welfare providers. (4) Only organisations that formalise, first, ownership and, second, control over their operations as private are considered.3 This criterion is intended to distinguish welfare associations from public organisations. First, the owners have to be private persons or organisations, not public authorities, and, second, the influence of public authorities should be limited: associations are not generally obliged to follow orders coming from public authorities, but need to negotiate with

5 associations if they want to gain influence. Public authorities may, for example, hold single seats on boards or influence decisions such as client uptakes or veto rights in exchange for public recognition or transfers. However, as previously stated, such public influence must be limited. In sum, the range of organisations considered in the study is limited to organisations that share a functional purpose (they aim to produce welfare in a regulated, profit-limited way) and the privacy of their ownership and governance (they are not a part of public administration and are relatively independent in the pursuit of “producing welfare”). In conjunction with this conceptualisation, one could ask: Who are the “private” owners of welfare associations? What motivations do they have to “aim to produce welfare?” How do they accomplish “welfare production?”

Assumptions: gains and resources The owners of associations are multifarious individuals and social groups, such as employers, churches, cooperatives, mutual societies, foundations, philanthropic associations, pedagogical or ideological organisations, et cetera. The widespread activity of associations across various social segments indicates a broad range of motivations for engaging in welfare provision. As described in associations’ own announcements, motivations range from self-help in addressing members’ welfare concerns, to altruistic-philanthropic motivations to help clients, to the furtherance of more general ideals or beliefs in combination with welfare production (Lorentzen 1994, James 1987, see also this thesis: day-care section 4).4 Moreover, the range of their activities stretches over various areas, including poverty relief, provision of social services, social insurance, schools, housing, and so on. Besides such specific motivations, a number of more general motivations to achieve basic gains through welfare provision may be assumed to underlie various associations’ welfare engagement across social segments. 1. One basic gain for all the associations, for example, could be their interest to control which type of welfare is produced in order to target specific client groups and satisfy the demands of supporters. Associations are active in a wide range of welfare areas, such as poverty relief, social services, social insurance, schools, housing, and

6 so on, but it is seldom that a single association is active in all of these areas. By engaging in welfare operations themselves, association founders can ensure that the welfare concerns and interests of their clients and supporters are met, for example, by defining the welfare area, setting benefit levels, and defining client access regulations. Furthermore, associations that operate welfare facilities can decide on operational aspects, such as connections to and influence over external grantors, personnel hiring, and so on. 2. An additional basic gain could be that welfare activity, like every organisational activity, implies the establishment of an organisational infrastructure and its associated human and monetary resources. By means of these, welfare associations can intensify their social interactions and communications. As welfare associations, they have the opportunity to interact with staff, members, clients and the wider society. Associations can communicate via meetings, informational material, national media, and so on. Furthermore, with the ability to make its own recruitment decisions, an agency can select personnel or clients based on patronage considerations.5 In sum, by being active as organisations, welfare associations can enhance social connections • and communication channels. 3. A further gain may be derived from welfare research. In welfare research it is often assumed that the provision of welfare enhances the social prestige of the welfare producing agency. The connection between social prestige and welfare is embedded in the idea that welfare provision appeals to a broad segment of the population and that the providing welfare agency, and especially its elites, can gain approval and acceptance through the provision of welfare (Rimlinger 1971, Flora and Alber 1995; Alber 1987, Mishra 1977; Skocpol 1992). A welfare gain by associations based on a desire to achieve social prestige is not part of every associational engagement. Very small-scale self-help groups, for example, may merely attempt to solve their own welfare problems. By contrast, all associations that have been established in a “top-down” manner by initiative takers not intending to be clients could eventually be motivated by the possibility of enhanced social prestige. Examples of top-down methods of establishment are associations founded by charitable individuals or by large social organisations, such as business firms, unions, and churches. In sum, welfare associations are regarded to have “welfare production” as their purpose. They are assumed to want to control

7 welfare operations in order to ensure that their welfare concerns are met, in order to enhance social interaction and communication, or to be perceived as socially prestigious. Irrespective of the degree of emphasis placed on social communication and prestige, such general gains from engaging in welfare activity are here assumed to underlie the associational welfare engagements that span multi-faceted social origins and welfare areas. Welfare associations share not only “privacy of ownership and governance,” but also a “purpose to produce welfare.” In conjunction with this, one could ask a third question: how do welfare associations fulfil their “purpose to produce welfare?” In order to produce welfare, associations need resources, be they in the form of human capacities or monetary means. The resources can come from various sources, such as public authorities, clients, volunteers or donors. The source of public authorities is of particular importance in funding associational welfare and this will be elaborated upon in subsequent sections. The dependency on resources, however, confronts welfare associations with the expectations and demands of resource providers. Public authorities, for example, often demand the appointment of board members in exchange for transfers, or they demand that associations use specific sets of statutes that are legally acknowledged (e.g., charities, cooperatives, et cetera). In other words, associations’ principle of independence will eventually be compromised by their need to satisfy external resource providers (associations’ resource situation will be discussed later in this introduction). In conclusion to this brief introduction to welfare associations, one could assume that associations are private, independent welfare pursuers that are confronted with the wishes of external resource providers. Due to their recurrent reliance on transfers from public authorities, the duality of privacy and dependency on public acknowledgement places associations at the interface between politics and social mobilisation. In this regard, associations share the attributes of unions in corporatist arrangements or political parties, yet, thus far, they have received relatively little attention in comparative welfare state research.

8 2. Earlier research on the preconditions of associational welfare provision As will be investigated later, the significance of associational welfare provision varies greatly between different welfare states and also between different areas within individual welfare states. It is one of the main aims of this study to understand the causes of these variations and to focus particularly on associational welfare in the two Scandinavian countries Sweden and Norway. Earlier research has focused on the role of the state, the role of certain socio-political power balances and the role of societal heterogeneity for associational welfare provision. Regarding the first aspect, there is a significant debate between two research camps: those who presuppose a negative and conflicting relationship between statutory welfare and associational welfare, and those who advocate the opposite view.

The role of the state Many researchers, in particular economists, claim that there is a trade- off between statutory welfare and associational welfare provision. • This perspective has most prominently been advocated by Burton Weisbrod and Estelle James (Weisbrod 1977, James 1987). Weisbrod and James assume that those welfare benefits that cannot be produced by the market are normally provided by the state, i.e. if the state has the capacity and the political backing to do so. If this is the case, and if the state thus satisfies citizens’ welfare demands, there is no demand for, and hence no room for, associational welfare provision. In other words, under these conditions associational welfare provision is “crowded out” by statutory welfare provision. The argument concludes that there is only room for associational welfare if the state fails to satisfy citizens’ welfare demands (state failure). Thus, Weisbrod and James not only claim that statutory welfare is an obstacle to the establishment of associational welfare, but also inherently assume that states and associations will act independently from each other as welfare providers, if they engage in welfare provision. Both, the state and associations, will provide welfare as well as financers, regulators and operators separately, without any exchanges. This implies that Weisbrod and James would also assume that associations have the capacity to act as significant welfare providers on their own – without support from the state.

9 Researchers of this strand, moreover, often advocate a conflict perspective in order to depict associations’ motivation to provide welfare (James 1987, 1989a). With respect to the interests of associational elites, James explains that associational welfare would often be a competitive strategy used by religious entrepreneurs to attract individuals towards their religion. In her view, religious entrepreneurs “use the non-profit form because their objective is to maximise faith or adherents rather than monetary profits. They consequently choose to enter those industries (such as education and health) that foster taste formation, socialisation, and group identification.” She expects associations to be “concentrated in geographic areas with strong independent religious groups competing for clients.” The examples she provides are associational schools, which were predominant in the religiously heterogeneous Netherlands. The assumption of a conflict motivation among associational providers could also be interpreted as indirectly advocating a conflict perspective concerning the state-association relationship. Based on a zero-sum assumption, statutory welfare activity is perceived as if it is in conflict with associational elites’ welfare activity; hence, associational elites and state elites are assumed to have fundamentally conflicting interests when it comes to the question of associational welfare. Associational elites are assumed to prefer to provide welfare on their own, whereas state elites are assumed to prefer public welfare solutions. The assumption of a trade-off or a conflicting relationship between statutory welfare and associational welfare has been strongly questioned in more recent research, which argues that state activity is a positive precondition in favour of rather than an obstacle to associational welfare. Advocates of this perspective claim that associations are not capable of developing a significant role as welfare providers if they are not receiving statutory financial support (e.g., Salamon and Anheier 1996, Kendall and Knapp 1996, Kuhnle and Selle 1992). They refer to empirical findings that indicate that public financing is an important revenue source for associational welfare. Based on this view, the Norwegian social scientists Stein Kuhnle and Per Selle have developed the notion of a partnership between the state and welfare associations. For Norway, they argue that states and associations did not perceive each other as adversaries, but as equal parts of society (Kuhnle and Selle 1992). In Norway, partnership and cooperation are apparent in several forms of interaction between

10 associations and the state. Examples are statutory subsidies for associations, associations’ participation in the legislative process as consultants in official committees and the relevance of broad popular movements for the development of the welfare state.6 In conclusion, this second perspective hypothesises that state activity in the welfare sector is a positive rather than a negative precondition of associational welfare, because statutory support is seen as an essential part of a successful associational welfare activity. Moreover, a partnership instead of a trade-off or conflict relationship between the state and welfare associations is assumed. In sum, departing from these two essentially different views on the role of states for associational welfare, different notions about the state-associations relationship have been formulated. The state- associations relationship is regarded as either being based on a trade- off or conflicting ideological interests of elites in associational versus public welfare solutions or being based on consensual, national partnerships between associations and states.

Historical contingency and the role of certain socio-political power balances • Between these two extreme positions concerning the state-associations relationship, we can identify a third position, which may be labelled the “contingency-position”. According to this third view, the actual relationship between statutory welfare and associational welfare is dependent on historical choices of welfare provision, which are, in turn, influenced by the balance of power between different socio- political forces. According to Salamon and Anheier’s social origins theory, the significance of associational welfare depends on the historical influence and interests of statutory elites and certain socio- political forces (Salamon and Anheier 1998).7 The relative historical predominance of different socio-political forces led to more or less associational provision in a country. According to Salomon and Anheier, the entities that are favourable for the growth of associational provision are: weak working-class movements that prefer statutory provided welfare, weak “landed interests”8 that oppose all forms of welfare provision, strong independent churches that are interested in their own welfare provision and a strong liberal bourgeoisie that prefers associational welfare provision to statutory welfare provision. Inspired by Esping-Andersen, Salamon and Anheier developed four models of third-sector regime types. Each model was described as

11 “reflecting a particular constellation of social forces:” a social democratic, a statist, a liberal and a corporatist model, respectively. Associational provision is supposed to be low in statist and social democratic regimes with predominant “landed interests” and working classes, and high in liberal and corporatist regimes with influential middle classes and churches. Examples of the social democratic regime type are Sweden and Italy, and of the statist type, Japan. Examples of the liberal regime type are Great Britain and the USA, and of the corporatist type, Germany and France. Salamon and Anheier remain unspecific about the historical phases within which they suppose the trajectory of state-association relations were determined. However, their reference to both “landed interests” and liberal middle classes as influential ”social forces” seems to suggest the beginning of industrialisation as the historical phase referred to. In conclusion, the level of associational provision in a country is seen as a result of the predominant power balance between the above- mentioned socio-political forces around, probably, the beginning of industrialisation. The historical compromises reached in these phases are regarded as likely to endure due to path-dependency processes. Therefore, in the end, historical socio-political power balances are seen as the crucial factor that determines whether associational welfare activity in a given state will be significant or not.

The role of ethnic and religious heterogeneity While disagreeing about the role of the state and the relationship between states and associations, most scholars hold rather similar views about the role of macro-structural preconditions. The main common denominator is that societies with a high level of ethnic and religious heterogeneity are regarded as more likely to develop significant associative welfare solutions than more homogeneous societies. Weisbrod assumes that associational welfare solutions are more likely to occur in heterogeneous countries. He supposes that societal heterogeneity undermines the chances for political consensus in relation to the introduction of public welfare measures. “State failure” is thus assumed to occur frequently in heterogeneous countries, which leads to unsatisfied welfare demands among minorities, whose members will then seek to satisfy their demands through associations. In homogeneous countries, on the other hand, necessary majorities to

12 support statutory welfare are more frequently achieved. Therefore, in homogeneous countries the state will be able to satisfy welfare demands, thereby crowding out associational welfare. Consequently, Weisbrod sees cultural, ethnic and religious heterogeneity as a basic positive factor behind associational welfare growth (Weisbrod 1990). James shares Weisbrod’s basic view, but her main argument is supply- rather than demand-sided. James interprets associational welfare provision as a strategy of religious entrepreneurs who wish to attract individuals towards their religion. Religious organisations “use the non-profit form because their objective is to maximise faith or adherents rather than monetary profits. They consequently choose to enter those industries (such as education and health) that foster taste formation, socialisation, and group identification.” She expects associations to be “concentrated in geographic areas with strong independent religious groups competing for clients.” The examples she provides are associational schools, which became predominant in the religiously heterogeneous Netherlands, while they disappeared in religiously homogeneous Sweden.

• Synthesis All in all, the perspectives that account for variations in the significance of associational welfare described above, give rise to three basic explanatory factors (see table 1): First, the state (or, to be more precise, statutory welfare activity) is assumed to play a major role. However, the character of the relationship between the state and associational welfare is disputed. Some researchers assume that intense statutory welfare activity is a negative precondition for associational welfare activity. Others suppose a positive or even a partnership relationship between statutory and associational welfare. Second, the historical power balance between certain socio- political groups is assumed to matter. Salamon, Anheier and other researchers seem to agree that a high level of influence of working- class elites leads to the monopolisation of welfare provision by the public sector and hence to a crippling of associational welfare activity, whereas the political dominance of religious and liberal-bourgeois elites is regarded as conducive to associational welfare solutions. Third, there is a strong consensus about the role of societal heterogeneity for the chances of an establishment of significant associational welfare. Religiously, culturally or ethnically

13 heterogeneous countries are expected to be more likely to develop significant associational welfare solutions than countries with a more homogeneous social structure.

Table 1: Background factors for associational welfare activity

Influence on associational welfare activity

Presumed background Positive Negative Irrelevant factors I. Statutory welfare Kuhnle/Selle, Weisbrod, James - activity Salomon/Anheier II. Influential working- - Salomon/Anheier Kuhnle/Selle class elites III. Social heterogeneity Weisbrod, James - Kuhnle/Selle

Norway and Sweden as challenging test cases The approaches described above are certainly helpful for understanding some of the variations of associational welfare activity across countries and welfare areas. However, the examples of associational welfare activity in the areas of childcare and housing in Norway and Sweden do pose a puzzling challenge. Firstly, according to one of the two approaches on the role of the state, Sweden and Norway, as typical Scandinavian countries with a high degree of statutory welfare involvement, should not be able to develop any significant associational welfare activity in a major welfare state area. Secondly, and also in accordance with the power-balances approach, both countries should be characterised by a low level of associational welfare provision, because of the traditionally strong socio-political role of working-class organisations in the form of social democratic parties and labour unions. Finally, the low level of ethnic, cultural and religious heterogeneity in these countries should also pose an obstacle to associational welfare activity. Yet, contrary to these accounts, a significant level of associational welfare provision was achieved in the welfare state areas of childcare and housing in Sweden and Norway. In fact, the level of associational welfare provision in these areas has exceeded that of some other Western European countries with less “welfare stateness,” less working-class strength and more societal heterogeneity.

14 As table 2 shows, in Norway almost half of the childcare institutions are in the hands of associational providers. The Norwegian figure is on a medium level, comparable to England and Germany, exceeding the level in France and only clearly below the level in the Netherlands. A similar picture emerges when we look at childcare enrolment figures. In Norway about 15% of full-time childcare is provided by associations. This figure is somewhat lower than in Germany and the Netherlands, but it is on par with England and far higher than in France. Irrespective of the comparison figures, it seems fair to state that the level of associational childcare activity in Norway is significant. When we look at the area of housing, both Norway and Sweden display a significant involvement of associational providers. In both countries, associational providers account for 16% of the overall housing tenure. This market share is exceeded only by that of the Netherlands and it is considerably higher than that of all other the countries in the comparison group. In conclusion, Norway shows a significant level of associational involvement in both childcare and housing and Sweden does so in the area of housing. These findings contradict assumptions that associational welfare provision is not • significant in the Scandinavian countries. Furthermore, the differences in welfare areas between Sweden and Norway are not reconcilable with the understanding that states and associations are either in a partnership or in a conflict/trade-off relation. It may very well be true that Sweden and Norway, in comparison with other countries, have opted more often for public provision in major welfare areas such as pensions, health and education. However, in the two welfare areas considered here, day-care and housing, a different picture emerges, and the picture does not fully comply with earlier accounts of associational welfare activity.

Table 2: Associational1 housing and day-care provision, 1990s, shares in %

England Sweden Norway Germany Netherland France Italy s Housing tenure 5 16 16 9 35 9 (0) Day-care institutions 45 13 42 51 84 2 Na

Day-care, enrolled 14 6 14 22 21 1 Na children2 Political Liberal Social Social Christian Christian Christian Christian predominance3 democratic democratic democratic democratic democratic democratic 16 Confessional Homogeneous Protestant Heterogeneous Homogeneous Roman- heterogeneity Catholic State-association n/a Popular movement support n/a n/a partnership for welfare state, Consultancy, Stat. subvention of assoc. 1 non-public, which includes associational providers as well as private enterprises 2 the share of full-time enrolled children (excluding family day-care) of non-public provision, share in cohorts; 1996 children 0-4, Sweden 0-6, France 0-5, 1997 Norway 3 according to Huber and Stephens 2001; France could also be classified as politically liberal, because no dominant Christian democratic party evolved in France Sources: housing: see Appendix A table A4; day-care: Rostgaard and Fridberg 1998; Norway NSD

The need for a refined analysis of associational welfare involvement The cases of Norwegian and Swedish childcare and housing show us that the above-mentioned approaches are in need of a number of refinements in order to better explain variations in associational welfare involvement. In my view, the three approaches would benefit from modification in the following ways: First, the state-role account does not require any modification per se. That is, I do not intend to dispute the basic hypothesis that state action matters for associational welfare provision. On the contrary, I regard statutory welfare activity as a crucial frame for associational welfare provision. I do, however, regard as unrealistic those accounts that claim that statutory welfare involvement is a negative precondition for associational welfare provision. Instead, I will argue that associational welfare usually only flourishes if associations are supported by the state. However, states neither consistently act as partners nor as challengers of associations, but attempt to subsidise and regulate associational welfare depending on associations’ comparative capacity advantages and associations’ socio-political • complementarity to ease the pressures placed on states. In other words, the relation of states to associations is understood as a “practical” welfare problem as well as a socio-political problem. Second, the power balances perspective is not regarded as problematic per se in this thesis. Power balances between different social groups, in particular if they are expressed in varying party political constellations, indeed do play a role in explaining variations of associational welfare activity. The problem with the power balances approach advocated by Salamon and Anheier is rather that it assumes general and constant ideological preferences of certain socio-political groups and political parties for either statutory or associational organisation (Salamon and Anheier 1998). On the one hand, bourgeois and religious groups are expected to prefer associational solutions, whereas on the other hand, labour and social democratic groups are expected to prefer statutory solutions. In this thesis such a generalised assumption will be questioned. Instead it will be supposed that both left-wing and right-wing parties may prefer associational welfare solutions or statutory solutions, depending on whether they hold a particular interest in supporting certain social groups in their potential role as associational welfare providers. Such an interest can be assumed to arise if certain parties share common ideologies or

17 common constituencies with certain social groups. One of the most interesting implications of this perspective is that socio-political constellations characterised by a dominance of working-class groups and parties can also, under certain conditions, be favourable for associational (instead of statutory) welfare provision. This is in particular the situation in the cases of Norwegian and Swedish housing discussed here. Third, in its classical form, the heterogeneity approach regards competition and conflict between ethnic, cultural or religious minorities to acquire the favour, sympathy and loyalty of individuals as a necessary precondition for significant development of associational welfare. As such a situation is assumed to appear only in heterogeneous countries, in the end associational welfare is assumed to positively co-vary with heterogeneity. In my view, various societal groups can be interested in offering associational welfare provision in ethnic, cultural or religious homogeneous countries. (1) Societal conflict and group competition is also apparent in homogeneous societies, for example between socio-economic groups. (2) Competition between different organisations for the favour of the same individuals, as supposed by James, is seen as only one possible situation prompting organisations to engage in associational welfare.9 Associational welfare can be interesting for social groups that do not necessarily attempt to attract the same individuals, but rather different categories of individuals, because they do not perceive all individuals as attractable. Furthermore, associational welfare can be interesting not only for competitive, expansionist social groups, but also for integrationist social groups by activating hitherto passive individuals and by regenerating and prolonging the support of hitherto active individuals. In order to include integrationist, labour-related and socio-economic interests as motivations for associational welfare, the notion here proposed is that particularistic interests per se can be regarded as motivations for associational welfare. In sum, the approach chosen regards the state-associations relationship (as formed by regulatory and financial relations) as an essential frame for associational welfare activity. Furthermore, how this relationship will look is, in turn, also regarded as dependent on the role played by political parties as important influencers of statutory welfare activity, i.e., political parties are seen as “transmission belts” between statutory welfare development and welfare associations.

18 Moreover, it is not solely the role of the state (and of political parties) in supporting associational welfare provision that is seen as important. The interests of various societal groups in engaging in associational welfare activity are also assumed to matter. Whether or not societal groups want to be active in the welfare area does not necessarily depend on whether these groups act in a heterogeneous or a homogeneous ethnic, cultural or religious context. In principle, associational welfare can be a rewarding and interesting activity for all groups in society (also for those in a homogeneous societal context), as it has the potential capacity to enhance these groups’ attractiveness to individuals. Based on these considerations, in the following I will develop a novel framework for the analysis of associational welfare involvement. Initially, I will address the role of the state with respect to welfare associations. For this purpose a broader discussion of statutory and associational welfare activity will follow. Thereafter, I will focus on the role of political parties.

•

19 3. Statutory and associational welfare activity Many scholars have regarded the relationship between the state and associations as a central aspect for understanding the development of associational welfare provision. Characterisations of this relationship range from a trade off or conflict situation (Weisbrod 1977, James 1987) to a partnership of state and associations (Kuhnle and Selle 1992). Instead of presuming that a negative or positive general relationship between a state and an association, I will regard a state and an association as organisations, that will consider the possibility of entering into a cooperation according to the advantages and disadvantages that the potential ’partner’ is offering them. In other words, if a state and an association accomplish a partnership or end up in a conflictual trade-off is regarded as an open question: a question that will be decided upon in a complex decision-finding process by both a state and an association.10 What a state and an association consider as advantages and disadvantages of cooperation is hereby regarded to be based primarily on considerations about respective organisational capacities and socio-political complementarity. They will have to consider organisational capacities, e.g., what kind of resources the cooperating partner can access and if the partner can satisfy welfare concerns more efficiently, or if the cooperating partner operates differently and is able to satisfy welfare concerns more effectively. However, they will also take into consideration, if the partner is socio-politically complementary, e.g., does the partner pursue similar welfare intentions, or is the partner an interesting organisation to share social prestige with when cooperating publicly?11 In order to provide a general idea of statutory and associational organisational capacities and socio-political interests, recurrently mentioned features of statutory and associational welfare efforts are collected from studies and reports. The aim of the compilation is to single out typical differences between ‘a state’ and ‘an association,’ which are, in turn, interpreted as incentives and disincentives for a state and an association to enter into a cooperation with each other. I will argue that a state and an association have different organisational capacities, which, firstly, provide them with access to qualitatively different resources and, secondly, enable them to qualitatively operate their welfare efforts differently. These different abilities are assumed

20 to make cooperation attractive for a state and an association, but especially for an association, as an association has less capability in accessing resources to access resources. Thirdly, it will be argued that a state and an association command over different control abilities, which can result in a limitation of independence for an association or a loss of control for a state in the case of cooperation. Their limited control abilities are here regarded as a disincentive for them to cooperate. Organisational capacities, however, are not the only criteria, that will influence an association’s and a state’s considerations. They will also consider if they can regard each other as socio-politically complementary. I will argue that a state and an association tend to differ, for example, concerning the client categories and welfare areas they choose to focus on. I interpret this as an indication of the tendencies of an association to pursue more particularistic welfare intentions, i.e. to address the specific welfare concerns of specific segments of the society. These tendencies are in contrast with a state’s more nationally-oriented intentions towards wider circles of the citizenry and their welfare concerns. Diverging welfare efforts and • intentions are, in turn, interpreted as indications of socio-political non- complementarity between a state and an association, which is assumed to be a disincentive for them to cooperate.

Organisational capacities: A state’s and an association’s incentives to cooperate: access to resources and operational organisation Students of associational welfare provision often emphasise that an association differs from a state in how it operates welfare facilities and how it can access resources. In the following, I will present major findings concerning both aspects of difference between statutory and associational welfare provision in order to argue that these differences mostly represent incentives for a state and an association to cooperate.

State’s and association’s access to resources A state and an association have quite different abilities to access resources, which, in turn, afford them with different abilities to provide welfare. A state can rely on an acceptance to enforce the payment of taxes and contributions to social funds, whereas a welfare association cannot rely on such an acceptance. Through the enforcement of payments, a state is enabled to ensure that a significant

21 proportion of the national GDP is used to finance welfare provision. In Western European countries, for example, public total tax revenues amounted to 40% of the GDP and public social expenditure amounted to 25% in 2003 (OECD Factbook 2007).12 The enforced inflow of resources also enables a state to insure social spending in times of client withdrawal or in times of economic recessions. Additional, one capability of a state is that it has easier, direct access to financial markets. It usually has a higher credit credibility rate and can directly access international financial markets. Thus, a state is enabled to borrow with lower interest rates than a welfare association (an aspect, which has been especially relevant for public loan provision in the area of housing until the liberalisation of financial markets, ILO 1930). An association is regarded to be able to attract voluntarily provided resources by, e.g. volunteers and donators; a capability that a state is regarded to have to a lower degree (Finlayson 1994).13 However, the relative weight of volunteer work and donations for associations' capability to provide welfare is subject to debate (see for example Salamon 1987, Salamon and Anheier 1996, Salamon et al. 1999, Knapp 1989). Usually, scholars of welfare developments have emphasised associations’ financial vulnerability. Salamon speaks of ‘associational failure,’ due to their “inability to create resource on a scale that is both adequate enough and reliable enough. […]…only when contributions are involuntary, as they are through taxation, are they therefore likely to be sufficient and consistent” (Salamon 1987). Likewise, Alber, when commenting on voluntary insurances, considered the lack of finances to be a main aspect for associational problems: “Voluntary systems could not survive, because they only disposed over limited resources …” (Alber 1987). However, the overall level of welfare activity by volunteering persons is quite impressive in many European countries (see Appendix A table A1). Volunteers provide work at the level of several thousands of full-time employees, which can be seen as an important contribution to the welfare provision of these countries. Nevertheless, when looking at the relevance of volunteering as a resource for associations, the level of volunteering seems not to be high enough to sustain the level of associational welfare production (see Appendix A table A2). This weakness of volunteering may be due to some typical features of volunteer activity. It has often been emphasised that the importance of volunteers lies in supplementing rather than in

22 substituting professional welfare, because, firstly, volunteers are predominantly untrained in the welfare area they are active in and, secondly, volunteering levels are regarded to be too low to substitute for paid welfare staff. Volunteers have limited spans of working time and usually work for no more than a couple of hours a week. Volunteering is therefore often situated within social services, where hourly work can more easily be arranged, e.g. shuttle services, visiting services, meetings, spare-time activities, organisation of daily life, personal care or the passing of children. Overall, volunteering has a low presence in core areas of welfare in most European countries, but is concentrated in specific welfare areas for marginalized or highly distressed persons, e.g., mentally or physically handicapped persons, homeless people and drug-addicted persons (see Appendix A table A1 and Archambault 1997; Kendall and Knapp 1996, Knapp 1989; Salamon and Anheier 1996, Salamon et al. 1999; Badelt 1989). Besides attracting volunteers, associations are regarded to be able to attract donations as a specific resource. This resource is, however, usually not available for public authorities. The overall level of donations to associations is, however, rather low and seems to be of • minor importance for the funding of associational welfare production (see Appendix A table A2). Furthermore, it is difficult to assess, in as much associations attract donations independently from statutory interference, due to often allowed tax redemptions for donations.14 In sum, associations’ ability to access resources by themselves seems to be delimited. Regularly, at least 45% of associational revenues are obtained via public subvention, and often they exceed 50% (see Appendix A table A 2). A few further examples may illustrate the welfare achievements of associations with only minor public assistance15 compared with larger scale public assistance. Associational voluntary insurances have usually been unable to provide coverage for more than minor parts of the population (see Appendix A table A3). The table in the appendix shows, firstly, the increases in coverage levels of voluntarily organised sickness and unemployment insurances in some European countries at the time of the first year of central state intervention, and, secondly, after longer times of statutory intervention, but before the introduction of compulsory welfare schemes. The different levels between the two time-points indicate that it was only after increased public assistance that associational coverage levels grew. The impression is that associational welfare provision could only have

23 grown from a 'marginal' level of provision into a 'substantial' level of provision with statutory help. Similarly, associational housing did not reach significant levels of housing provision before becoming subsidised by a state. Earliest censuses show that before statutory housing policies became extended, associations’ shares of national housing stocks were low (see Appendix A table A4). In Great Britain 0,6 % of the 1914 housing stock was led by associations, 0,8 % in Sweden 1933, 2,6% in Germany 1939 and around 6% in the Netherlands 1947.16 After post- 1945 Western European increases in public intervention into housing markets, associational housing stock shares began to increase considerably. Housing stock censuses show that associations could enlarge their shares in several countries. Associations have reached shares of 37% in the Netherlands, in Sweden 16% and Norway 14%, in Germany 9%, in France 8% and in Great Britain 5%. (The growth of housing associations in Great Britain did not start until after the Second World War. British housing policy in 1974 made associations liable for generous public transfers; Hills 1989, Department of Environment). In most countries, public policies provided advantageous conditions for associations. Several aspects can be mentioned. Loan subsidies and the provision of loans were regularly available for housing associations. Furthermore, access to building sites was eased for associations. Tax deductions also played a role in assisting associations, but are regarded to be of lesser importance.17 The overall picture across several welfare areas is that associations’ own capability to attract resources is delimited and that associations are dependent on statutory revenues for their level of welfare provision. Therefore, it is assumed that an association is ‘financial vulnerable’ and, in turn, that statutory subvention of associational welfare is regarded as a crucial factor for their level of welfare provision. Associations are therefore assumed to be interested in cooperating with a state in order to access further resources for their welfare activity. However, an association has a greater capibility to attract voluntary resources than a state does. Thus, it is here assumed that a state can be interested in cooperating with an association in order to share its access to voluntary resources and to lower the expense of statutory resources.

24 State’s and associations’ operational organisation A state and an association differ not only with respect to their ability to access resources, but also in terms of how they organise their welfare operation. Public administrations are hierarchically structured and operate and decide in accordance with legally formalised procedures and rules. Clients are entitled to get their case decided upon in a neutral, consistent and objective way by public administrations. The decisions taken by public administrations are, in turn, legally binding, which implies that clients have the right to take legal proceedings in order to overhaul an administrative decision (Rothstein 2003). Furthermore, public administrations employ professional personnel, who are trained for their tasks, work consistently over time and are obliged to make decisions and work according to formalised rules. Public administrations are centralised, which means that sub-units have demarcated responsibilities in relation to geographical areas. Whereas, for example, several associations are active within a single municipality, often only one public administration is active (in larger cities, there are eventually several public sub-units). In addition, • public administrations work within a formal line of hierarchy and accountability, whereby super-ordinate units can order the execution of tasks by sub-ordinate units. Such hierarchic and centralised decision structures ease the coordination of tasks and balancing of resources between different units.

Welfare associations are often characterised by students who operate informally with horizontal, network-like decision structures, which are dissimilar to the formal and hierarchic structures of public administrations. Associations’ reliance on volunteers to exert both administrative tasks and welfare tasks could be said to be one reason for this specific way of operation. Volunteers are seldom subjected to formalised work tasks: a circumstance that is assumed to be conducive for informal, intuitive and non-binding reactions between personnel and clients and to encourage higher degrees of informal social interactions between the personnel. In addition, a low degree of formalisation enables an association to react more flexibly than public administrations to new welfare demands and is often assumed to be a reason why an association could act as 'pioneers' of welfare production in some areas, i.e. provide welfare before the state

25 becomes engaged (Hatje 1999, Beveridge and Deakin 2000, ILO 1930). A further, often-mentioned, difference between associational and public welfare organisation is that associations tend to be organised in small, independent units, which seldom are formally subjected or accountable to super-ordinated organisations (Lorentzen 1994, Kendall and Knapp 1996). This provides a high degree of autonomy for a single association, which further increases its flexibility and ability to provide welfare in an informal, non-binding way. However, such features of associational welfare provision are not typical for all associations. Housing cooperatives or larger churches, for example, employ rather large standing bureaucratic organisations, that operate in a highly formal, professional and centralised manner, which makes them comparable to statutory organisation. In sum, a state usually provides welfare in a professional, legally formalised, standardised and coordinated way, which entitles clients to administrative decisions about their ‘cases.’ However, the high degree of formalisation and standardisation may fall short of meeting the welfare demands of clients. The demands of the clients may not ‘match’ the prescribed standardisation. An incentive for a state to cooperate with an association could therefore be the wish to indirectly address clients and welfare demands, which a state cannot reach with its form of welfare organisation.18 Furthermore, a state may opt to subsidise associational ‘pioneers’ in a new welfare area, where a state does not, as yet, want to engage itself. An association operates often with the help of volunteers in network-like organisational structures and provides welfare in an informal way, which allows for intuitive reactions, informal interactions and flexible responses to clients’ welfare demands. However, volunteers’ unprofessional welfare treatment may not always have matched the clients’ wishes for a professional treatment (Hollingsworth and Hollingsworth 1989, Lorentzen 1994). Furthermore, volunteers' unprofessional administration eventually led to economic problems or even insolvencies of welfare associations (Beveridge and Deakin 2000). Thus, associations are confronted with the limits of their specific organisational capacities, which may enhance their wish to cooperate with a state, for example, in order to unburden themselves of a number of administrative tasks by shifting responsibilities to public administrations (e.g., day-care associations in Sweden).

26 In summary, a state and an association can access qualitatively different resources for operating welfare and both can organise welfare qualitatively differently. These qualitative differences could be said to enable them to satisfy welfare concerns in different ways: a state can offer large-scale, standardised welfare programs staffed with professionals, which provides many citizens with welfare rights and professional treatment. An association, on the other hand, can offer small-scale, informal and flexible welfare, which can treat clients more personally. Because cooperation provides both with the opportunity to indirectly access different kinds of resources and to indirectly offer different forms of welfare, a state’s and an association’s capabilities concerning resources and operation are regarded as incentives for them to cooperate. However, cooperation not only opens advantageous opportunities for both, but can, likewise, imply disadvantages, e.g., in the form of a loss of control over welfare production to a cooperation partner, which is the next topic.

State’s and associations’ disincentives to cooperate: limited control • A welfare association is an independent and private organisation and is free to regulate its activities. It can independently decide to operate formally or informally, to select personnel and clients, to define its constitutional statutes, to set the standards for administrative procedures and to set the standards for welfare provision, such as benefit levels, opening hours, client fees, et cetera. However, an association is confronted with the regulative power of a state, which is acknowledged as a compulsory prescriber of regulations concerning welfare production within its territory. This can have various implications for how an association can organise its activities. A state can, for example, interfere with the statutes of an association, including limitations about the use or distribution of yields, e.g. nonprofit criteria, distribution according to cooperative principles (see overview in table 3). Other forms of regulations can concern the internal decision structures of associations: prescriptions about stakeholder influence, forms of appointments to the board of representatives, the form of accountability of the board, et cetera. Even whole sets of statutes can be regulated, e.g. a charity, a trust, a foundation, a mutual society or a cooperative,19 which deal with several aspects of the internal organisation of associations.20 A state can also regulate operational aspects of associational welfare

27 production like benefit levels, opening hours, professional staff per client ratios, and so on.

28 •

Table 3 Statutory intervention concerning welfare associations

Regulations Subsidies Aspect Related demands Related incentives Constitution - Application and registration Eligibility to tax deductions (e.g. VAT, income (associational (e.g. non-profit, charity) taxes), general organisation subsidies statutes) Welfare operation Administration - Professional personnel Eligibility to welfare related transfers (subsidies - Administrative standards and contracts) (accounting, reporting) - Public auditing - Public board representatives 29 Production - Public inspection and monitoring Eligibility to welfare related transfers (subsidies - Prescriptions about client rights, and contracts) client fees, professional treatment, etc. Coordination Coordination agency - Registration Eligibility to welfare related transfers (subsidies with public and contracts) representation Welfare production - Client waiting lists, coordinated planning

Influences Æ Æ Type Level of associational welfare provision of associational welfare provision

A state has thus several possibilities to interfere with an associational organisation, which can lead to a severe limitation of independence, whether it be in the form of an alternation of its entire organisation and/or an alternation of the type of welfare it produces. In Beveridge’s words, states are the “masters of money” for associations (Beveridge and Deakin 2000). Therefore, an association, which would dislike the regulative demands from a state, would regard them as disadvantages for entering into cooperation. A state tends to demand compliance to regulations in exchange for transfers. In other words, welfare associations that cooperate with a state are more comprehensively regulated than independent associations. Thus, an association will usually have to weigh regulative demands against the financial incentives that follow with cooperation. In contrast to a welfare association, a state can rely on its relatively superior regulative powers to insure its independence and to influence the welfare production of welfare associations. However, this may not lead to a state’s perception that cooperation with an association secures its intensions with welfare production. Although able to regulate associational welfare provision in detail, a state is confronted with the ‘limits of regulation:’ the so-called ‘implementation problem’ (Sannerstedt 2001). When leaving operational control to another agency, the problem for a state is that the agency could operate contrary to statutory intentions, because the operating agency is able to more or less eagerly pursue public regulations and ‘intentions’ for its operation in its day-to-day decisions. The provider can, for example, influence personnel appointments and client-uptakes via personal interactions. Furthermore, the operating agency can, in informal ways of disapproval and persuasion, influence its personnel and clients, which could alternate the welfare produced. In sum, the providing agency has some leeway to influence day-to-day-decisions through direct contact with personnel and clients. The implementing agency will act according to the intentions of the decision taker, but only under the condition that it is willing (Lundqvist 1987, quoted by Sannerstedt 2001).21 A state that cooperates with an association will be confronted with the same implementation problem, i.e. an association’s possible evasiveness from official regulations. Only if a welfare association is willing to comply with official regulations, can a state be sure that its intentions are followed through. In other words, a state’s ability to

30 control associational welfare provision is limited, although it can issue official regulations. In sum, a state has delimited abilities to control associational welfare production. Such a loss could be perceived as a disincentive for a state to cooperate with an association; especially if it expects that its cooperation partner would divert from its intentions for welfare production. A state will thus have to weigh the possible advantages of cooperating with an associaiton, e.g., its access to volunteers or its informal operation, against the loss of control that comes with leaving day-to-day operation to an association. So far the focus has been on statutory and associational respective organisational capacities, which have been assumed to have influenced their considerations about cooperating with each other for welfare production. However, welfare-related decisions by a state and an association are not only considered to be based on considerations about organisational advantages and disadvantages, but also on considerations inasmuch a cooperation would further their socio- political intentions with producing welfare.

• Diverging welfare intentions of a state and an association In order to provide a general notion concerning associational and statutory welfare efforts, frequently mentioned features are compiled in relation to, firstly, associational welfare efforts and, secondly, statutory interference with associational welfare efforts. It will be argued that associations have often pursued specified provision of welfare, whereas a state tended to regulate in order to broaden it. These tendencies to diverge concerning their focus on welfare provision are interpreted as indications of diverging socio-political intentions with welfare production, which can lead to tensions between the cooperation partners and even to failures to agree upon the terms of cooperation. A welfare association’s pursuit of particular interests is regarded as an indication of the often observed variation of welfare goals and welfare production between single associations (Salamon 1987, Beveridge and Deakin 2000, Kuhnle and Selle 1992).22 A number of examples may illustrate such variations between associational providers. Associations differ, for example, concerning their officially proclaimed goals for welfare engagement. Whereas cooperatives

31 proclaim to provide welfare according to the interests of their members, philanthropic individuals state that their aim is to help indigent or disfavoured persons and denominational organisations aim to advance their beliefs connected to welfare provision (Fischer 1995, Lorentzen 1994, James 1989a). Furthermore, welfare provision by associations varied. Associations had, for example, different geographical areas as their main activity areas. Associations tended to concentrate their activities, e.g., in urban areas or geo-cultural areas where associations were able to address particular client groups. Housing cooperatives and employers, for example, tended to be active in industrialised areas in order to address the welfare concerns of blue-collar workers. Social service providers, however, tended to be active in urban service-sector dominated areas in order to address middle class clients. Denominational associations tended to be active in areas where adherents to their denominations lived (Beveridge and Deakin 2000, Lorentzen 1994, Gustafson 1974, Strömberg and Pooley 1992, Umrath 1953, Salamon 1987, see also this thesis: day-care section 6). Furthermore, associations focused their activities on certain welfare areas and neglected other welfare areas in order to address the welfare needs of 'their' clienteles. Churches and trade unions, for example, tended to be active in different welfare areas. Trade unions focused on founding unemployment insurances, sickness insurances or housing cooperatives, whereas churches tended to provide charity, education or to found hospitals (Beveridge and Deakin 2000, Salamon 1987, Lorentzen 1994, Flora and Alber 1995). The selection of clients by associations is another typical feature of associational welfare production. Philanthropic associations and churches, for example, often selected their clientele among the very poor and marginalised strata of society, whereas associations that provided retirement homes or housing selected clients that they considered to be 'pauvres honteux,' e.g. widows of former state employees or other professional occupations. Associational insurance schemes, which are often founded by various occupational groups, selected their clients exclusively from ‘their’ members, i.e. they selected clients from certain, specified occupations and professions. Risk-prone client categories such as female employees, unemployed persons or persons working in dangerous occupations, were not covered by, for example, associational sickness, pension or unemployment insurances (Beveridge and Deakin 2000).

32 Associational insurance schemes selected clients according to a horizontal occupation differentiation, but also according to a vertical risk differentiation, which often led to the exclusion of socio- economically weaker groups. A further feature of associational welfare provision was that associations shaped their activities according to their preferences in particular cultural norms (Power 1993, Salamon 1987, van Laarhofen et al. 1990). Schools organised by Churches, for example, emphasised religious education in their curricula (Green 1990, Good and Teller 1969). Likewise, religious associations providing day-care facilities or youth centres tended to design their welfare performance according to their convictions by, e.g. emphasising religious contents and selecting personnel (Lorentzen 1994, James 1987, see day-care sections 4 and 6). In addition, when dealing with welfare clients, associations exerted varying degrees of ‘paternalistic’ social control (Salamon 1987, Strömberg 1992, Holmans 1987). In the area of housing, philanthropists and employers, for example, exerted social control in the form of visits and inspections (e.g. Octavia Hill’s philanthropic • housing, employer housing in the Ruhr-area; Umrath 1953). French housing associations, for example, exerted formal social control by concierges; an institution they invented in France in the 19th (Power 1993). Other forms of social control by associations were directed at their employed personnel. Denominationally organised welfare associations, e.g., obliged their personnel to attend on prayers (Lorentzen 1994). In sum, associations selected among clients, welfare areas, geographical areas, worldviews or forms of social control, which led to variations between associational welfare provisions. The variations of welfare designs by associations are interpreted as indicative of a variety of interests to provide welfare in order to address specific welfare concerns and to be as attractive as possible to certain client groups. Associations’ multifarious owners and social backgrounds may here be seen as a reason for the variations of interests between single associations. In contrast, a state is assumed to pursue less particular welfare interests than an association or, more nationally-oriented welfare interests. Such nationally-oriented intentions of a state are put forward as indications of a tendency to regulate associational welfare provision in order to ‘de-particularise’ associations’ welfare provision.

33 A selection of examples for such tendencies of states to regulate associational welfare activity can be mentioned. A main tendency of statutory intervention concerning associations was the extension of client rights and coverage levels (see also table 3). As already described above, despite associations’ multifarious welfare efforts, associations’ only covered minor parts of national welfare needs and demands. States addressed this lack of coverage by extending client rights towards insurance providers in order to include client categories that were not covered. This happened, for example, by decreeing earmarked subsidies for the inclusion of riskier client categories like family members in sickness insurance, handicapped children in day-care, and so on. Other regulations with the aim to enhance coverage levels were the decreeing of maximum levels for client contributions and fees, in order to provide incentives for less affluent strata of the population to subscribe to social insurance or to attend social services (Beveridge and Deakin 2000, Perrin 1969, Flora and Alber 1995, Alber 1987, Lindqvist 1990). Statutory welfare intervention led, furthermore, to a more evenly spread welfare provision across welfare areas and geographical areas. States addressed these coverage biases by, for example, providing statutory earmarked subsidies for specified welfare services across the territory or by subsidising associational welfare provision across different welfare areas (Kuhnle and Selle 1992). Another set of public regulations concerned with the even spread of welfare provision were invoked to enhance coordination between associations through the creation of coordination agencies under public tutelage (Salamon 1987, Kendall and Knapp 1996, Beveridge and Deakin 2000, Perrin 1969). Recurrently, states tended also to impose standards for welfare operation in exchange for subsidies in order to ensure that the 'type of welfare' produced by associations was in accordance with statutory welfare interests (Beveridge and Deakin 2000, Lorentzen 1994, Kramer 1989, Lipsky and Smith 1993, Judge and Smith 1989). An example of such a standard was the introduction of subventions for day-care providers, who had to open for certain minimum time-spans or to provide specific personnel/children ratios (see day-care section 3). Other examples are regulations that address social control efforts or cultural settings by associations, e.g., by limiting the extent of religious teaching in Church-run schools or by ensuring the privacy

34 rights of applicants and clients concerning associational providers (see also day-care section 3 and housing section 5). In order to address the possible problem of associations' evasion from statutory rules, states, for example, created semi-public agencies under public tutelage (Kendall and Knapp 1996, Beveridge and Deakin 2000, Perrin 1969). Furthermore, associations usually had to accept public representatives on their boards and public monitoring in exchange for public subsidies (Lipsky and Smith 1993). Moreover, states tended to impose standards for administration, e.g. professional staff, in order to secure economic and legal soundness (Beveridge and Deakin 2000). In order to ensure a targeted use of statutory subventions according to states' interests, states have repeatedly created legal types or ‘sets of statutes,’ that standardise associational organisation. Regulations itemised, for example, limitations about the use or distribution of yields (e.g. non-profit criteria, co-operative distribution), prescriptions over internal decision structures (e.g. democratic type of stakeholder influence, appointment of board representatives, accountability of the board) or whole sets of statutes, comprised to several regulations (e.g. British Charities, German • Gemeinnuetzige Vereine, Scandinavian Housing Cooperatives). States also tried to indirectly balance the use of resources between associations through, e.g. the pooling of risks by the prescription of an extension of access rights for clients or by performance related contracts, which tightly regulated subsidies for performed services (Beveridge and Deakin 2000, Kendall and Knapp 1996, see day-care section 3 and housing section 1). In sum, states addressed associational provision in various ways. Statutory interference recurrently ‘broadened’ welfare activities, which resulted in enhanced coverage levels across various social categories, welfare areas and geographical areas. Often, such decrees had a socio-economic dimension, e.g. decrees concerning the rights of riskier or less affluent persons to be accepted by an insurance society, decrees concerning maximum levels for contributions and fees, or decrees of earmarked subsidies for the inclusion of specific less affluent persons or for certain performance standards. It can also be argued that states likewise have delimited associational provision, e.g. by standardising eligible client categories, by setting welfare production standards and by coordinating associational welfare provision, which may eventually have resulted in excluding certain

35 clients. However, the overall impression of the data suggests that statutory intervention led to an extension of welfare provision. Statutory regulations of associational welfare provisions are interpreted as attempts to impose more nationally oriented statutory welfare intentions onto associations’ particular welfare operation.23 Where associations were regarded as ‘too narrowly’ following their particular concerns with some client categories, states tended to regulate associations’ operations in order to broaden it socially and to encompass various welfare concerns.24 The interpretation promoted for a state’s differing socio-political intentions from an association’s is that a state is exposed to demands and concerns from all the mobilised and organised citizens in its territory, including, for example, demands for welfare provision of union members and church members on the one hand and taxpayer interests on the other. In contrast, an association is confronted by the pressures of its membership and support basis, which is comprised of only a specific segment of the population.25 Following this understanding of welfare production, statutory welfare efforts are interpreted as the attempt to balance and reconcile national pressures for both effective and efficient welfare production, while an association is assumed to be more focused on the pressures from the citizens of a specific segment to (most effectively) satisfy their specific welfare demands.

Summary The aforementioned compilation of typical differences between ‘a state’ and ‘an association’ concerning their organisational capacities and welfare intentions was meant to provide an idea of the recurrent advantages and disadvantages for both to cooperate for welfare production. A state and an association have been characterised as organisations with different levels of capabilities to access resources and to organise welfare operation, which have been interpreted as implying mostly incentives for them to cooperate. Further capacity differences between a state and an association are their regulation powers. An association eventually has to accept regulations by a state in exchange for transfers, which could infringe its independence and thus be perceived as a disincentive to cooperate. In contrast, a state is confronted with its limited ability to control an association’s welfare

36 operation in detail, which likewise could imply a disincentive for it to cooperate. Besides considering organisational capacities for the question of cooperation, a state and an association are assumed to consider if their cooperation partner is socio-politically complementary, e.g. inasmuch as the cooperating agency pursues similar welfare efforts or whether the cooperating agency is regarded as an interesting organisation to share social prestige with. Reported tendencies of a state and an association to differ in their welfare provision are hereby interpreted as indications of differing welfare intentions and, in turn, socio- political non-complementarity. These differing socio-political interests are assumed to weigh as disadvantages for a state and an association when deliberating over cooperation. In conclusion, it is assumed that a state and an association will decide upon cooperation in a complex decision-finding process whereby they will weigh possible incentives and disincentives resulting from their respective capabilities and socio-political intentions. Associations, for example, can regard states’ capability to access • resources via compulsory taxation as a strong incentive to cooperate with states, but associations are often confronted by regulative obligations that come with the eligibility to statutory transfers. Whether or not associations regard statutory regulations as an incentive or disincentive to cooperate, depends on whether or not associations regard regulations to be in conflict with their own 'particularistic' welfare intentions. For example, associations were regularly only able to meet the welfare concerns of more affluent strata (thus coming close to private market provision). If they would have been eager to extend their provision concerning less affluent strata, they may have regarded statutory regulation in exchange for enhanced resources to be an acceptable ill. In rare cases, associations, with welfare concerns from lower strata, may even have preferred statutory welfare provision, because they regarded their own provision to be insufficient (e.g., in some cases in Scandinavia, Kuhnle and Selle 1992; see also day-care section 4). Furthermore, if regulations would be in line with their own interests and if they would regard themselves as not capable to enforce similar regulations, due to their relatively weaker regulative powers, associations could regard states' capability to regulate their activities as an asset. This was, for example, the case in Norway as housing cooperatives urged the state

37 to introduce certain tenure conversion restrictions (see housing section 4). States could regard cooperation with associations as advantageous, because associations have capabilities to access certain types of resources, i.e. volunteers and donations, which are to a lesser degree accessible for states.26 States could also have regarded associations’ informal welfare organisation as an incentive to support them in order to indirectly address welfare demands in a more personal way or in order to indirectly be able to react flexibly on welfare demands. Nevertheless, states are confronted with the burden of control and its limits. If states would see a need for evermore detailed prescriptions and monitoring of associations, associational employment could not only pose an enormous control burden for them, but, in the end, would eventually still fail to ensure that statutory intentions would be followed through, i.e. if associations would be reluctant to oblige. This could be the case, if associations would continue to follow through their own 'particularistic' welfare interests irrespective of statutory regulations. For example, associations could disregard regulations concerning the privacy rights of applicants (see day-care section 5). However, if associations could provide welfare at lower costs than public administrations states could eventually still accept a narrower 'particularistic' welfare provision by associations, because of pressures to provide welfare more efficiently. It is argued that the complexity of states' and associations' welfare decisions are further increased by their respective deliberations concerning how advantageous or disadvantageous cooperation would be for states’ and associations’ wider socio-political interests in welfare production, e.g. to safeguard or to share 'social prestige.’ If states were to perceive associations as connected to socio-political groups pursuing unwelcome ideas and interests, they could regard the gain of social prestige by associations to be a problem. The preclusion of denominational schools from public subvention in Turkey, Italy, France or, partly, Germany could be regarded as an example of such a case of socio-political non-complementarity. Thus far, the focus has been on statutory and associational welfare activities. In the following, the focus is narrowed down to the state, because it is regarded to be the agency that decisively influences its own and associational welfare provision concerning level and type of welfare production, based on its capacity to prescribe compulsory tax payments and to prescribe binding welfare regulations.

38 The leading question will also focus on the state: under what circumstances does a state cooperate with a welfare association? Statutory action is assumed by mainstream welfare research to be influenced by its organisational capacities, by socio-economic processes, which result in functional welfare needs, or by socio- political processes of interest pressures (see note 11). Although the main approaches in welfare research do not explicitly address questions about how welfare is organised, one could adopt the notions of these approaches to questions concerning statutory welfare decisions about cooperation with associations: Firstly, if states choose associations according to whether they can cover the needs of the population more effectively, or whether they can efficiently keep statutory expenses low, state decisions to employ associations could be seen as determined by either statutory capacities and/or functional welfare necessities. Secondly, if states choose associations because they provide welfare shaped in accordance with particularistic interests and therefore cover more effectively the welfare interests of important socio-political groups, state decisions to employ associations could be seen as determined by socio-political • pressures. Finally, states could cooperate with associations as they would regard associations to be socio-politically complementary and especially 'worthy' of the social prestige' of welfare provision. If one were to consider the variations of statutory employment of associations in Western Europe, one could argue that, to some degree, all aspects could have influenced statutory decisions. The recurring pattern of, for example, the public employment of associations in social services in combination with less prevalent associational involvement in the area of large social insurances in Western Europe, could be understood to be based on associations’ relative capacity advantages to attract volunteers for social service production, which makes associational social service production more efficient in those countries. However, the same variation could also be dependent on socio-political influences by Churches and related Christian Democratic parties on statutory welfare policies, because most of the associational social service production in Europe is operated by Church-related welfare associations (e.g. Caritas, Diakonia). Therefore, it is difficult to distinguish between the influences of the various factors, especially without a more thorough empirical investigation.

39 In this paper, I will focus on one approach; the so-called power resources approach. According to this approach, welfare state action can mostly be regarded as a response to socio-political pressures from various social organisations pursuing their specific, particularistic interests, e.g. economic interest groups, welfare associations or political parties. I think that the emphasis on the influence of particularistic interests on statutory action is an interesting angle for attempting to understand statutory policies towards associations. Welfare associations (and related clients, members, supporters) can be assumed to be intrested in safeguarding their specific welfare provision. But how are particularistic interests in associations’ welfare production carried into the political decision process? Of course, welfare associations can act as pressure groups themselves in order to influence statutory decisions concerning financial and operational regulations. However, among the various social organisations that attempt to influence statutory welfare policies out of particularistic interests, political parties, not welfare associations, are understood to be the organisations that have a crucial position in the political decision-making process. A subsequent question is then, how can the interest of political parties in welfare associations be understood? The interconnected themes of the next section are political parties’ positions concerning statutory policies and their interest in welfare associations.

4. Political parties Political parties are understood as the principle organisations that organise social interests to control elective statutory institutions in order to secure influence over public policies (cp. the definition of LaPalombara and Anderson 1992, Ware 1996). In contrast with other private organisations such as welfare associations, political parties have an immediate influence on public policies. This is not to say that parties are not themselves influenced by statutory administrations or pressure groups, but that other organisations have to convince parties to account for their interests. Parties, for example, decide on the selection of personnel, who will occupy the elective positions of statutory institutions. Furthermore, political parties are those organisations that discipline parliamentarians to vote according to partisan interests. They also pre-select policies before these are decided upon in government and parliament. Positioned as they are in

40 the political process, political parties are assumed to be of crucial importance for influencing statutory decisions in general. As such, they also play a major role in the funding and regulation of associational welfare. But how can political parties’ interest in associational welfare be understood? Even though they occupy a special position in relation to statutory institutions, parties are not statutory organisations as public administrations, but private organisations, that depend on social support from certain citizens and certain social groups. Therefore, parties will organise social interests in a strategic way. They will, for example, represent the demands from certain citizens and social groups, but will likewise mediate, arbitrate and even invent policy demands from and for their constituencies (Mair, Plasser and Müller 2004a). Therefore, as political parties are oriented in relation to the interests of constituencies, which are not comprised of the whole citizenry, they are here assumed to act as particularistic organisations. Parties’ interests in welfare associations are, in turn, assumed to be influenced by their orientation towards particularistic interests. Associations may pursue welfare provision that does or does not, as • far as parties are concerned, serve the welfare interests of their constituencies. As a result, parties will pursue policies, that either favour or disfavour associational welfare production. The policies pursued will depend on whether a party considers the form of associational welfare to be in its own interests or not. However, this is not to say that parties will pursue public policies out of particularistic interests in a narrow sense. In order to insure a long-term electoral success among the majority of voters, who are welfare clients, taxpayers and employees, they will, for example, have to account for which type of welfare provision will most efficiently and effectively address national welfare problems. This may also include the acceptance of already established associational solutions, because a radical alternation would be too costly. Furthermore, governing parties are also influenced by statutory administrations, which will remind them about those general statutory interests described in the section above. Thus, parties are assumed to be influenced by both raison d’etat considerations and particularistic interests whenever they conduct welfare politics. The assumption that partisan interests in welfare associations vary and will lead to a pursuit of differing policies is shared by a number of researchers, such as advocates of the power resources approach and

41 Salamon and Anheier. However, I argue that their stance in relation to the reasons that parties favour or disfavour associational welfare is too narrowly focused on ideological stances towards more or less generous welfare provision. Advocates of the power resources approach, for example, tend to focus on the “colouring” of the government by political parties and assume that the partisan “colouring” of the government leads to different constructions of welfare states. Depending on whether a government is left-wing, conservative or right-wing, welfare states will be more or less likely to take on a large-scale and universal structure. No explicit statements have been made by the prime advocates of the power resources approach about the influence of socio-political power balances on associational welfare provision. However, they seem to argue that left-wing governments pursue welfare policies that aim at the equal, universal access of all citizens to welfare treatments. If one holds the position that public administrations underlie to a greater degree obligations to follow public regulations of a bureaucratic ‘blindness’ towards the case of a client, one could presume that this implies a statutory organisation of welfare production. This would imply a preclusion of a statutory employment of ‘particularistic’ welfare associations and they may tend to follow their own interest in specific client categories. Therefore, one could interpret the power resources approach as implying a left-wing government preference for statutory welfare organisation, which would, in turn, imply the avoidance of associational organisations. In other words, one could be led to understand that advocates of the power resources approach would presume that left-wing parties are ‘non-particularistic’ concerning the question of welfare organisation. In a similar fashion, third-sector theorists tend to assume that social democratic countries do not favour associational welfare provision, whereas social conservative and liberal countries are generally expected to favour it. Salamon and Anheier, for example, derive partisan interests from ideological stances to a “strong state,” which, in the case of social democratic parties, leads to a preference for high government welfare spending and statutory welfare organisation in order to insure a “strong state” (Salamon and Anheier 1998).27 One could interpret their reasoning in terms of an indication that a social democratic interest in generous welfare provision, which would be a

42 similar form of reasoning to that of the advocates of the power resources approach. In sum, third sector theorists and advocates of the power resources approach seem to make the assumption that working-class parties in general take a negative stance towards associational welfare, whereas bourgeois parties take a positive stance. Based on these considerations it could be derived that they suppose that left-wing parties choose not to support any significant autonomous associational welfare provision and that parties of the political right would be more supportive. This perspective will be challenged here. In this thesis it is assumed that political parties are not generally interested in either supporting or not supporting associational welfare, as supposed by Salamon, Anheier and others. Instead, it is hypothesised that, in principle, all parties, irrespective of their political colouring, can be supporters of associational welfare. Yet, they will do so in different ways and each with a view to their particular interests in supporting certain welfare associations with certain welfare provisions in certain areas. Political parties will act as supporters of association friendly policies if they share an intersecting and • overlapping socio-political basis (or “constituency”) with the welfare associations seeking statutory support. This is, for instance, the case when social democratic parties support the welfare efforts of working- class-related housing associations, or when Christian democratic parties choose to support the welfare activities of churches or religious organisations. Political parties may also act as opponents to association friendly policies if they regard the social basis of the welfare association seeking statutory support to be 'incongruent' with their own. The reasons for a perceived incompatibility can be that the social basis supports political opponents, pursues opposing goals or belongs to opposing networks. This perspective on the role of political parties might be labelled the “social-congruency” hypothesis, as it claims that political parties in charge of government will pursue an associational–welfare-un- /friendly policy whenever they are confronted with associational welfare efforts by organisations that have social constituencies that are “in-“/“congruent” with their own constituency. Political parties can be expected to pursue such strategies as the opposition/support of in- /congruent organisations and their welfare activities (which are mostly targeted at their own constituencies) may indirectly help them to stabilise and integrate the opposing parties’/their own constituencies.

43 This, in turn, might help them hinder/ensure long-term stability and loyalty of theirs/their voter reservoir. The advantage of the social congruency hypothesis is that it makes it possible to understand those cases where working-class parties support associational welfare as well as those where right-wing parties choose not to do so. In classical associational welfare research, which is based on the assumption of a generally negative stance of working- class parties and the opposite stance of right-wing parties, these cases would have posed a puzzle that cannot be resolved. The perspective that I advocate rejects the idea that certain political parties hold fixed general positions towards associational welfare, and attempts to resolve this puzzle. A further difference between my proposed assumption and former approaches is that the interests of parties in associational welfare are not exclusively based on party stances towards more or less welfare provision, but that parties can be interested in welfare associations’ capabilities and functions for various reasons. Welfare associations can, for example, influence their welfare provision to address the welfare concerns of client categories, which can also be interesting for parties. In addition to their main task, welfare associations can provide clients with information about political decisions and highlight the role of a specific political party concerning welfare-related decisions. Welfare association can also acquaint clients with normative stances of, for example, organising social interactions (quasi-market, cooperative, corporatist, et cetera) and norms of social conduct (individualistic, solidaristic, status-oriented, et cetera). Thus, welfare associations may further partisan interests to actively influence socio- political socialisation. A close connection between a welfare association and a political party would also open for both opportunities for vice versa elite education or patronage. Theories of mass party mobilisations as formulated by students as Kirchheimer, Duverger or Panebianco served as a background to the social congruency theory. They observed that mass parties, especially socialist and denominational parties, were often involved in networks with auxiliary organisations (Kirchheimer 1966, Duverger 1967, Panebianco 1988). While these theorists developed their ideas on the examples of mass party involvement in socially rather closed subcultures, the idea I propose assumes that parties can also be interested in welfare associations that are less closely connected to a party. In a general sense, one could formulate partisan interest in

44 welfare associations to be based on welfare associations’ specific functions, which lie beyond partisan capabilities, and which, thus, indirectly help party organisations to widen parties’ social reach towards constituencies and, in rarer cases, to enhance their access to resources. In conclusion, the social congruency thesis proposes that political parties pursue divergent policies concerning welfare associations. These divergent policies are based on their different interests in more or less socially congruent associations. As a result, partisan advocacy concerning statutory policies towards associations is assumed to lead to conflicts between parties. However, this is not to say that governmental or state actions are driven only by these particularistic interests. Governing political parties cannot be regarded as fully autonomous actors, but their influence is limited in various ways: by other parties, by powerful interest groups, by welfare associations’ engagement and even by their own will to take political risks and to account for raison d’etat considerations. Therefore, political parties will often consensually agree upon statutory policies concerning associational welfare • production, irrespective of their own particularistic interests. However, political parties will, at the very least, attempt to influence regulations according to their own interests in some cases of associational welfare production. With respect to welfare policy, this implies that they will follow a strategy to support welfare associations and activities in those areas where they see a chance to further this interest, whereas they can be expected to oppose or to retard statutory support for associational welfare activities that would serve the interests of their political contenders.

The hypothesis of this paper is that statutory support for welfare associations is moulded by political parties in accordance with their particularistic interests towards welfare associations. Political parties will try to influence both statutory subvention and the regulation of welfare associations and, thus, attempt to influence both the level of associations' welfare production and the type of welfare that associations provide. Furthermore, political parties will pursue conflicting policies concerning specific associations in variation with partisan interest in associations with either more or less 'congruent constituencies.'

45 In this study, political parties’ influence on the development of associational welfare provision in the areas of day-care and housing in Sweden and Norway is analysed. Emphasis will be placed on party politics and its variations as they relate to associational welfare provision. Main questions are, first, whether political parties differ in their attempts to influence statutory policy formulation concerning associational welfare provision, or whether they unanimously decide upon statutory policies. The appearance of partisan conflicts would indicate partisan interests in associational welfare provision, whereas partisan consensus would indicate parties acting according to a raison d’etat. A second main question is whether parties pursue policies concerning associational welfare provision that conflict with those of congruent constituencies. Indications of congruency between a supportive party and a supported association include, for example, sharing of policy goals, overlapping memberships and coinciding supportive segments of the population.

The structure of this study After this introduction, the two main chapters on day-care and housing will follow. Each chapter is divided into seven sections:

• Section 1 introduces the reader to the respective welfare areas. It describes the functions of welfare provision and related historical initiatives by associations in order to provide a general notion of associational interests in day- care. This serves as background information for later sections, which describe conflicts between parties over statutory support for associational welfare provision. • Section 2 examines the extent to which welfare associations have successfully provided welfare in their respective areas independent of statutory assistance. This is undertaken to examine empirically the theoretical assumption made above, that associations are not capable of significant welfare provision without statutory assistance. How independent are associations from statutory support and partisan assistance? • Section 3 provides an overview of statutory intervention, focusing on regulations and subventions that could have

46 affected associational provision. It serves foremost as background information on the complex body of related statutory policies for the following chapters on partisan policies and associational variety.

In sections 4 to 6 the social congruency hypothesis is empirically examined. • Section 4 looks at engaged associational providers in greater detail. The aim is to gather information about the variety of active associations in an area. It considers the question of how helpful the shaping of statutory subventions and regulations has been for the success of the involved associations. • Section 5 describes the political process. The focus here is on the interests of single political parties concerning policies for associational providers. Which parties advocated and which opposed policies favouring associations? • Section 6 compares the social connections and orientations • of main advocating parties and main engaged associations, examining whether they have coinciding interests in welfare provision and in organising overlapping constituencies. • Section 7 provides a brief summary of the main findings.

Following the two main chapters on day-care and housing, in the conclusion I will summarise the development of associational welfare and discuss the results in the context of the theoretical perspectives presented earlier in this introduction.

47 48 II. Day-care in Sweden and Norway

In present-day Norway and Sweden, associations are involved to a considerable degree in day-care provision. This degree of involvement is similar to most Western European countries (see table 2). However, this has not always been the case. Whereas Norwegian associations were engaged in day-care provision on the same level as public providers since the Second World War, Swedish associations nearly vanished from the child-care market during the 1960s and 70s and made a late recovery since the 1980s. How can these differing developments be understood? It will be argued that the differing developments in Sweden and Norway have been strongly influenced by public subvention and regulation of associational day-care, because associations had to rely on statutory resources to expand their day-care provision (section 2). However, public subsidies cannot be regarded as the exclusive factor for the differing levels of associational engagement. Associations’ own interests in being active (under publicly set regulations) seem to • point towards another differentiating factor between Sweden and Norway (sections 3 and 4). Furthermore, it will be argued that the enactment of more or less favourable statutory conditions for associational welfare was influenced by partisan politics in accordance with their interests in the mainly involved associations in Norway and Sweden (sections 5 and 6). Before I turn to addressing associational welfare in Sweden and Norway, I will begin by providing an idea of the interests in day-care provision.

1. Introduction: Day-care The leading question for this introduction is which interests do persons or organisations have in founding and owning associational day-care facilities? In general, persons and organisations are assumed to be interested in owning a welfare association in order to control its operation.28 As owners, they are regarded to have the opportunity to immediately and in detail shape the operation, whereas other persons or groups, for

49 example, clients or financiers, financiers can only apply pressure to promote their interests indirectly. Thus, persons and organisations that wish to control the operation of day-care facilities are assumed to be interested in owning welfare associations. The question arises which day-care specific operations could persons or organisations have an interest to control? To answer this question I will provide an overview of the functional variations resulting from different operational modes of day-care provision in order to provide an idea of how differently day- care provision can be configured. I will then mention a few ‘extreme’ historical examples of associational day-care operation in order to provide a notion over how associations diverged when they shaped their day-care operations towards differing functions. Diverging functions are hereby regarded to be a result of the diverging interests of associational owners. Finally, I will mention further aspects, which owners could have an interest in to control.

Historical development of functions and associational day-care provision Child day-care is a social service delivered through direct personal contact between grown-up personnel and child clients. As a result of the personal contact and the age difference between personnel and clients, day-care personnel can be expected to have some form of socialising influence on their clients. Besides this basic function of day-care, day-care can fulfil further functions for its clients. These functions depend on the chosen specification and combination of operational items, such as length of opening hours, attending children categories, employed personnel, choice of curricula and choice of pedagogical methods. In order to provide an overview of the large variety of possible operational modes and functional outcomes, I would like to emphasise three sets of operational modes and resulting functions that I think have been influential since the beginning of modern day-care provision, namely the function of educating children in certain areas of knowledge, of pedagogically developing children and of supervising children. The function of educating children could be said to be predominant for day-care facilities that operate in a school-like manner. In practice, a day-care personnel in such facilities is advised to lecture children and to require them to concentrate on what the teacher is saying. To insure that children are able and concentrated enough to follow the

50 taught curriculum, only older children are encouraged to attend and this for only a few hours per week. Furthermore, the owners of educating facilities would exercise considerable care when, for example, selecting and instructing their personnel. They would require of applicants a specific education in or engagement for a specific curriculum, which they would be expected to teach. Examples of day-care facilities that place great emphasis on educating children are Sunday schools. The Catholic Church, for example, was providing day-care in the from of Sunday Schools on a broad basis in France and Belgium at the beginning of the 20th century, but Sunday schools could also be found in other European countries run by protestant denominations (Green 1990, SOU 1951:15). Bible lessons or teaching of the Catechism were the main occupations of such denominational day-care. In later periods, the labour movement also organised Sunday schools (for example in Germany and Scandinavia). Second, facilities with a pedagogical function are characterised by the usage of planned activities designed to encourage children to activate themselves. To instruct children according to pedagogical • methods, the employment of pedagogically trained personnel would be of importance. Furthermore, such facilities would attempt to provide stimulating environments for children by, for example, providing opportunities to play with specifically designed toys or to be outside in a natural environment. In consideration of the children’s socio-emotional ability, pedagogically oriented facilities are opened only shorter hours and accept only older children. The use of pedagogical methods and techniques to optimise the individual development of children can be understood as a scientification of child-raising, whereby pedagogical advice was seen as an important supplement to “natural” upbringing. Pedagogical ideas and methods influenced the development of day-care facilities at a relatively late stage. Typical facilities with an emphasis on pedagogical ideas and methods are kindergartens, as developed by Fröbel, Montessori and others. It was only in the 19th and 20th century that the Kindergarten movement spread to all western countries; it was supported by upper- and middle-class women (Ds U 1985:5, Hatje 1995, Antman 1996). A third function, which was predominant for the operation of some day-care facilities, was to supervise children. Work tasks for the personnel of such facilities mainly included the guarding of large

51 groups of children, but not necessarily their intellectual stimulation or education. Children of any age could attend such facilities, because children can be guarded irrespectively of their age group or their state of mind, e.g. while at sleep. Such facilities offered longer opening hours than facilities with a pedagogical or educating function. Furthermore, the employment of trained personnel would be of lesser importance for supervising facilities, because untrained personnel could also supervise children. Supervisory facilities have their origin in modern urbanisation and industrialisation. At this time the working and living places of parents became separated, thus isolating children from working parents. Unsupervised, often criminal children became a new social concern in urban areas, which led to the spread of facilities for child supervision during the 20th century in Europe. The aim of these early centres was the supervision of children in a rather “raw” form with opening hours into the late evening and often very large groups of children. Typical facilities were crèches, day-care centres or asylums,29 mainly provided for single mothers and mothers in economically weak families. Philanthropic groups or churches founded such facilities out of a desire to provide social assistance, whereas employers of female- dominated industries, such as the textile industry, provided day-care with long opening hours for their female employees (Hatje 1995, Ds U 1985:5, NOU 1972:39). This situation changed in the last third of the 20th century, as increasing numbers of women entered the labour market, which now included even middle-class women (see Appendix B figures A6 and A7). The function of supervision, which was formerly stigmatised as social assistance for less-well-off women, became connected to female demands for economic independence. Opening hours and accessibility of day-care throughout the week, typically to also fulfil the function of supervision, became a broader concern for day-care (Rostgaard and Fridberg 1998, Antman 1996). The focus on these three functions can, of course, only be regarded as a very simplified picture of the functional and operational nuances of day-care provision. Pedagogical methods, for example, are today used in most forms of day-care provision. However, the here highlighted functions are still assumed to provide a rough orientation concerning current traits of day-care variation. In figure 1, functions and related forms of associational day-care facilities are depicted.

52 In addition to the elaborated functions that have been discussed, which were specifically related to day-care operation, further functions that are also common for other welfare areas can be thought of as possible variations. Day-care facilities can, for example, be shaped in order to address specific segments of society. Through the combination of quality standards and fee requirements, for example, day-care facilities can address specific socio-economic segments. To address ethnic segments, for example, day-care facilities could provide day- care with an emphasis on using the languages of respective ethnic segments. In sum, day-care can be provided to fulfil various functions. Due to the variability of day-care provision, persons, who hold an interest in one or the other function, may aspire to become owners of day-care facilities, because, as owners, they are assumed to have a pre-emptive influence on shaping the operation of day-care facilities. Based on this view, one could interpret the development of associational day-care into various branches to be a result of associations’ diverging interests in day-care provision. For example, employers in textile industries or in care-related branches, which • depend on female employees, would establish day-care facilities in order to provide day-care with work-related opening hours for their work force. Likewise, churches that wish to acquaint children with their religion would found their own day-care facilities in order to insure that curricula include bible contents and that pre-school teachers are adherents of their religion. Furthermore, owners interested in catering for well-off parents would become involved in day-care provision in order to offer high quality day-care with high staff-children ratios and exceptionally well-trained personnel with above-average parent fees. However, although various associations became active as day-care providers, associational day-care was usually limited without public aid.

53 (EMPLOYERS OF, E.G., THE TEXTILE INDUSTRY, SERVICE INDUSTRY, SCANDINAVIAN UNIONS, GROUPS OF THE LIBERAL/LABOUR WOMEN MOVEMENT) E.G. ASYLUMS, CRÉCHES, EMPLOYER DAY-CARE FACILITIES

WORKING TIME RELATED OPENING HOURS, CHILDREN OF ALL AGES

SUPERVISION

EDUCATION PEDAGOGY

TEACHING OF WORLDVIEWS SHORT OPENING HOURS, OLDER CHILDREN USE OF PEDAGOGICAL METHODS

E.G. SUNDAY SCHOOLS E.G. KINDERGARTENS (CHURCHES, WELTANSCHAUUNG GROUPS) (PEDAGOGICAL GROUPS, SUCH AS OF THE KINDERGARTEN MOVEMENT)

Figure 1: Functions, facilities and associations and interest groups [FUNCTIONS, ‘TYPICAL’ DAY-CARE PRACTICES, EXAMPLES, (INTEREST GROUPS)]

2. Associational limitations and statutory intervention This section examines the extent to which Swedish and Norwegian associations have been capable of providing day-care without public assistance. In order to estimate associations’ capabilities, associational provision is traced back to the time before and after the introduction of central state subvention. Minor associational provision before statutory intervention and associational expansion after statutory intervention is assumed to indicate that associations indeed have been interested in providing day-care to a larger extent, but have not been capable to do so without statutory assistance. Before the first national intervention into day-care in Sweden and Norway, day-care was dominated by associational provision. More than 90% of all day-care facilities were associational in Sweden 1943 and, in Norway, the share of associational facilities was above 80% of all facilities in 1960 (St.meld.nr.89 1961-62). Nevertheless, the total number of day-care facilities provided by associations remained limited before national intervention in both countries. In Sweden, 342 • associational facilities existed in 1943 before the central state started to subsidise day-care, and in Norway, 259 facilities existed in 1960 just three years before the first issuing of statutory grants (see Appendix A tables A7 and A9). Therefore, day-care provision dominated by associations remained on a level that could be called marginal. This changed with the introduction of statutory subsidies. In Sweden, the first national subsidies were introduced in 1944 and were followed by several subsequent schemes until the 1960s, but the overall expenditure level remained modest (see Appendix A table A8) However, even these low levels of subvention increased the number of day-care facilities, which trebled from 342 to 1170 until 1960 (see figure 2 and Appendix A table A7). The main expansions of central state intervention into day-care followed after 1963 and 1976 respectively, when central state expenditure for day-care was raised from 7 million Swedish crowns in 1963 to 1245 in 1976 (SFS 1963: 379, 1966:172, 1973:1205, 1207 ff.). The rise of subvention levels was followed by steep rises in the number of day-care facilities, from 1170 in 1960 to 4324 in 1970. In Norway, the central state introduced day-care subsidies for the first time in 1963 in form of permanent state grants for operational costs. In 1966, building grants for day-care facilities became 55 additionally provided (NOU 1972: 39). The introduction of these early and delimited day-care schemes by the central state was already followed by a considerable increase in the number of day-care facilities from 259 in 1960 to 883 in 1975. The main day-care expansion phase started with the Day-Care Act of 1975, when central state expenditure increased substantially (see Appendix A table A10). Since this decisive enhancement, day-care provision increased explicitly with the trebling of the number of day-care facilities from 883 in 1975 to 2384 in 1980. The expansion continued under subsequent Day-Care Acts in the 1980s and 90s and the number of facilities reached 4648 in 1990 and 5830 in 2000 (see figure 2 and Appendix A table A9).

16000

14000

12000

10000

8000

6000

day-care facilities, counts 4000

2000

0 1943 1960 1970* 1980 1990 1995*

Sw eden Norw ay

Figure 2: Day-care facilities in Norway and Sweden, counts * Norway: 1975, 2000 Source: see Appendix A tables A7 and A9

In sum, statutory intervention was followed by a considerable expansion of day-care provision in both countries (see figure 2). But how did statutory intervention influence associational provision? Have associations been eligible to subsidies and did associations participate in the growth of day-care provision? In Sweden, associations were eligible for national subsidies from 1944. During the first decades of eligibility until 1960, associations increased the number of their facilities. Afterwards, however,

56 associational day-care facilities diminished, although associations were still eligible for statutory grants. Due to the increase of day-care facilities by other providers, the share of associational day-care decreased to only 3% in 1970 and less than 1% in 1980 (see figure 3).30 By contrast, associations enhanced their provision considerably during the 1980s in Sweden. Since the beginning of the 1980s, associations increased the number of their facilities to 1 986 in 1990 and 2 354 in 1995. They could even increase their share of day-care provision from less than 1% in 1980 to 19% in 1995 (see figure 3). Norwegian associations increased the number of their facilities in direct reaction to state funding. Already from 1960 to 1975, when the share of associational provision dropped, associations doubled the number of their facilities from 219 to 415. Afterwards, associations further increased the number of their facilities and could even uphold their relatively high share of day-care provision amid the expansion of overall day-care provision (see figure 3).

100

• 90 80

share ofnon-public facilities, % 20

0 1943 1960 1970* 1980 1990 1995*

Sw eden Norw ay

Figure 3: Non-public day-care facilities, shares in % * Norway: 1975, 2000 Source: see Appendix A tables A7 and A9

In sum, Swedish and Norwegian associations did not provide day-care beyond a marginal level before central state intervention. Only after the central state introduced day-care subsidies did day-care provision

57 expand, moving from marginal associational provision to substantial associational provision. This indicates that Norwegian associations and later also Swedish associations were eager to expand their provision with statutory support. Therefore, they can be assumed to have been otherwise incapable of providing day-care on the same level and that they depend on statutory resources for upholding their later-achieved level of day-care provision. However, in Sweden during the first decades of eligibility to statutory subsidies associations were reluctant to expand their provision. Their reluctance could either indicate that they were not interested in expanding their day-care provision per se or that they were not interested in applying for subsidies due to the conditions posed to them in exchange for statutory subsidies. In order to understand the different developments of associational involvement, the following sections will investigate whether the different developments of associational day-care in Norway and Sweden can be understood in terms of associations' varying degrees of interest to provide day-care themselves (section 4), and whether the conditions of statutory intervention were less favourable for associations in Sweden than in Norway (section 3).

3. Statutory intervention and associational provision Statutory intervention into day-care mainly affected associational day- care by setting the conditions if and how they can access public funding. Therefore, two sets of regulations are concentrated on: (1) central state regulations concerning the eligibility of associations for public funding and (2) central state regulations concerning day-care operation.31 (1) Funding regulations are here regarded as crucial for associational day-care provision, because they can have implications for associations both for their absolute level of day-care provision and for their relative competitiveness with other providers. Therefore, the focus on funding regulations will not only be on how associational providers are addressed, but also how regulations are formulated for associational providers in comparison with other providers. Different forms of funding will be considered. Foremost, the focus will be on regulations for basic operational subsidies, such as grants for personnel expenses or grants for building expenses, which have been 58 in force since the introduction of state subsidies in both countries. In addition, regulations of time-limited subsidies or subsidies for special purposes will be considered, and especially if such regulations differentiated between associations and other providers. Furthermore, regulations by the central state will not only be considered with regard to central state subsidies, but also with regard to municipal subsidies. Municipalities have been important financiers of day-care in Sweden and Norway, because municipal subvention amounted to around 30-50% of day-care operators’ expenses (Blom- Hansen 1998). Usually, municipalities could relatively freely decide which providers they would like to support. However, central state regulations eventually enforced eligibility of associations to municipal subsidies, thus, considerably enhancing associations’ rights to public subvention. Due to their relevance for associations, such regulations will also be considered. (2) Besides funding regulations, regulations concerning operational standards can also have implications for associational provision. Operational regulations, for example, that prescribe high quality standards such as lengthy opening hours, services for younger • children, and educated personnel, could have negative effects on associational provision because they are cost-intensive and demand a certain standard of managerial capability from associations that they may not possess. Under the condition of tight subsidies, high quality standards could have negative effects on associational provision in comparison with municipal providers.32 Likewise, regulations that circumscribe the level of client fees that providers are allowed to demand, could limit associations’ ability to raise further resources and thus threaten the existence of associations when public subsidies are scarce.33 Furthermore, operational regulations that specify curricula, personnel selection, opening hours, the age groups of children or client fees could counteract associations’ own ideas and interests in shaping their operations. For example, regulations that specify curricula could infringe on some associations’ interests in teaching worldviews or using certain pedagogical methods. Likewise, operational regulations, such as the setting of maximum levels for client fees, could limit associations’ aspiration to provide exceptionally high-quality day-care for more affluent clients. Therefore, statutory regulations can influence associational provision in various ways, ranging from associations’ economic

59 situation to associations’ freedom to shape the operation of their facilities according to their interests. Below, the development of how central states regulated associational day-care in Sweden and Norway is described. Leading questions for the description will be: Have associations been eligible to public funding? Have they received public funding under the same conditions as other providers? Left public operational regulations associations elbowroom to follow their own ideas and interests?

Sweden From their first issue in 1944 until the 1970s of statutory day-care funding, national day-care schemes made no explicit differences between provider categories in Sweden. Central state funding, which predominantly subsidised salary costs of day-care providers, was available for all provider categories, including associations. Likewise, conditions for receiving public funding were the same for all providers both concerning official legal regulations and their specifications by the National Board of Health and Welfare (Socialstyrelsen).34 However, even though associations were eligible under the same conditions as other providers, associational facilities disappeared almost completely amid a quadrupling of all day-care facilities during the 1960s. As a result, the share of associational provision dropped from 38% in 1960 to 3% in 1970 (see figure 3 and Appendix A table A7). A possible explanation for this negative development of associational day-care provision could lie in the decrease and later abolishment of public funding of short-time day-care, because many of the then active short-time facilities were owned by associations (Ds S 1965:3, SOU 1972:27). During the 1940s and 1950s, a place in short-time day-care (Kindergarten)35 was subsidised with around 20% of the amount paid for a place in day-care with half-and full-time day-care.36 This was changed at the beginning of the 1960s, when public funding for short- time day-care became less generous. Then, the subsidy paid for a place in kindergartens was lowered to 10% in 1962 and 8% in 1963 of the amount paid for a place in longer time day-care.37 Finally, in 1966, public funding for short-time day-care was abolished. Only facilities, which have at least two thirds of the children for 5 hours a day, were still eligible to statutory funding. Thus, facilities, which exclusively

60 provided day-care for short-time attendance, lost their former eligibility. In addition, subsidies for constructions and renovations of day-care facilities, which were introduced in 1963 and prolonged until 1976, were exclusively granted to half- and full-time day-care, but not to kindergartens (Socialstyrelsen Råd och Anvisningar nr. 163, 1963 SFS 1963:397, SFS 1966:173, SFS 1973:1207).38 In consideration of the growth of expenses for building and operation subsidies in the 1960s, the preclusion of kindergartens from both subvention and building subsidies can be regarded as a considerable disadvantage for this form of day-care, resulting in the relatively slowed growth of short-time day-care during the second half of the 1960ies (SOU 1972:27). Therefore, also associational day-care providers, who are often engaged in providing short-time day-care, were confronted with considerably changed conditions, which may have contributed to their decreased activity in the 1960s.39 In the 1970s, conditions for associations became even more precarious. Municipalities were required to provide day-care for six year old children and to develop plans about future demands for day- • care to identify the increases needed in 1973. However, associational provisions were decreed non-equivalent to municipal provisions inadmissible towards fulfilment of municipal obligations (SFS 1973:1205, Prop. 1973:136). After 1976, associations had to negotiate for statutory funding with municipalities instead, whereas municipalities were free to decide whether or not to provide the requested assistance as long as associations fitted into municipal day- care plans (SFS 1976:381, SFS 1976:396). This further deterioration of publicly set conditions for associations can be regarded to have resulted in the almost complete disappearance of associations as day- care providers until 1980. Furthermore, associations’ eligibility for statutory grants was delimited concerning funding for short-time facilities. Statutory subsidies were re-introduced for short-time day-care, thus including all forms of day-care activity in 1973. However, associational providers were only eligible for subsidies for regular full- or half-time day-care, but not for short-time day-care, which were their main day- care activities until the 1970ies (SFS 1973:1207). During the 1980ies, governments started to issue “positive lists” about the categories of associations, which were eligible to statutory funding. The first category of associations to gain eligibility in 1983

61 was those ‘with a certain ideal or pedagogical orientation.’ Thereafter followed parent cooperatives in 1984 (Prop. 1983/84:9; SFS 1983:943; 1984:552). Associational providers, who remained excluded until the beginning of the 1990ies, were for example employers, personnel cooperatives, housing foundations and the Swedish State Church (Prop 1990/91:38, SFS 1990: 1188, SFS 1991:237, SOSFS 1992:1). Associations were also precluded from extra personnel grants, which were exclusively given to public day-care providers until 1987 (Prop. 1983/84: 9; SFS 1983:943; 1984:552; 1987:860). Additionally, associations were precluded from funding for short-time facilities (SOSFS 1991:3, 1989:47). At the beginning of the 1990s, all delimitations to associational eligibility were finally abolished. From 1992, all provider types could receive subsidies (also for-profit providers). In addition, non-public providers became eligible for subsidies for all forms of subsidised day-care activity, including short-time day-care (Prop. 1991/92:65, SOSFS 1992:1).40 In 1994, associations even became eligible for subsidy levels that were equal to those granted to municipal facilities, thus establishing for the first time associational access not only to statutory but also to municipal subsidies (Prop. 1993/4:11).41 After the introduction of these legal regulations, associations can be regarded as having achieved full equality with public and other providers. Although associations had in several aspects restricted eligibility for public funding during the 1980s, associational day-care provision started to grow. In contrast to the decades up to the 1970s, the number of associational facilities rose and associations even enlarged their share of day-care provision. The institutional changes in the 1990s were followed by a further increase in associational day-care (see figure 3). Furthermore, the increase of associational day-care came irrespective of the rather tight operational regulations for day-care providers, which have been in force since the 1970s in Sweden. By comparison, requirements for pedagogically trained personnel, personnel density and the size of children’s groups are high (see Appendix A tables A11-13). Associations had apparently no further problem to adjust to these requirements. They provide day-care to similar standards as public providers concerning, for example, the sizes of children’s groups, levels of professionalisation and opening hours (Skolverket 2005:261).

62 Not even the introduction of a maximum limit on parent fees for child-care in 2002, called maxtaxa, seems to have been a problem for associational providers. Until 2002, parent fees were only recommended “not to inhibit parents from applying for day-care” and to be “proportional to parents’ incomes.” These were non-binding recommendations, whereas the maxtaxa prescribed how to configure parent fees.42 The maxtaxa had to be applied by all providers, but was expected to pose greater difficulties for associational providers, because it was assumed that associations had demanded higher fees from parents than their municipal counterparts. Contrary to expectations, associations adjusted smoothly to the new regulation and, as more recent studies show, they demanded comparable fees from parents at parity with public providers (Råd och anvisningar 1963, nr.163, Socialstyrelsen anser 1978, nr.2, Skolverket 2003:231, 2005:261).

Norway In contrast to Swedish associations, Norwegian associations participated from the beginning of statutory day-care intervention in • the growth of day-care provision. Furthermore, during the main expansion phase after 1975, associations increased the number of their facilities and the share of associational day-care remained at a high level despite an overall increase of facilities (see Appendix A table A9). It is rather impressive how quickly associations reacted to the extension of public funding: within a decade they were able to expand their day-care facilities fourfold. The question arises: have regulations in Norway differed in such a way from those in Sweden that it can be argued that they have alleviated the participation of associations? In Norway, state grants were available for associations under the same conditions as for public providers under all day-care schemes since 1963 (for-profit providers remained excluded). Furthermore, funding for short-time day-care had not been abolished, but remained generous by comparison with long-time day- care (Blom-Hansen 1998). In 1980, for example, day-care for up to 3 hours was financed up to 30% of its real costs, whereas day-care for more than 3 hours was financed only up to 20% (St.meld.nr. 93 1980- 81). This may have supported associational day-care provision, because associations seldom offered day-care for younger children and long opening hours until the beginning of the 1990s (see figure 4).

63 40

care, 35

provider, % 15

fulltime day-carefulltime for younger day- children of all 0 1975 1980 1986 1988 1990 1992 1995 2000

public 40+, 0-3 associational 40+, 0-3

Figure 4: Opening hours and age of children by public and associational provision, shares in % Source: NSD

At the beginning of the 1980s, the central state ameliorated funding for operating day-care facilities with longer opening hours and for younger children in comparison with short-time day-care for older children. In addition, extra subsidies were issued by the central state to transform day-care for older children and short opening hours into day-care for younger children and long opening-hours (see Appendix A table A15).43 These changes of funding did not, however, lead to a decrease in associational provision. After 1986, associations increasingly began to offer longer opening hours and care for younger children. In consideration of the fact that funding for short-time day- care for older children remained available, associations’ extension of opening hours and the admission of younger children can be understood to be an indication that short-time day-care for older children was not necessarily an interest of associations per se, but may have been a choice due to insufficient public funding for long-time day-care (an underlying reason for the rise of associational long time provision and provision for younger children was an alternated composition of associational providers, which will be elaborated in the next section over associations).44 Further regulations that could have been more favourable for associational participation in Norway than in Sweden, are regulations 64 concerning operational control, which can curb associational independence, and regulations concerning the quality of provision, which imply high costs for the operators of day-care facilities. Operational control of associational day-care remained limited in Norway compared to Sweden. For example, associations did not have to fit into municipal provision plans nor did they have to incorporate their own waiting lists into municipal waiting lists, as Sweden did (Day Care Acts 1975, 1995). Instead, associations could make their own decisions on matters such as opening hours and client placements irrespective of municipal demands for certain forms of day-care. In addition to greater independence form municipal day-care, particular associations gained increased independence from general statutory day-care regulations at the beginning of the 1980s. Originally, the Day-Care Act of 1975 obliged all day-care facilities “to develop their activity in close connection to children’s parents.” This was amended in 1982 and associations gained the right to direct their activities towards a clear religious orientation, irrespective of parents’ wishes and legal regulations to adopt a moderate Christian orientation (Ot.prp.nr. 75 1981-2, Day-Care Act 1995).45 • Further independence was given to particular religious associations concerning their personnel recruitment. Such associations were freed from legal restrictions that forbade employers from asking applicants about their personal convictions and beliefs (§55A Arbeidsmiljöloven). Instead, certain associational providers gained the right to ask such questions and choose applicants based on their responses (Ot.prp.nr. 75 1981-82). Besides having more independence concerning their operation in Norway than in Sweden, associations had to meet lower quality requirements for statutory subsidies with regard to standards for personnel density, the sizes of children’s group and professional staff (see Appendix A tables A11-13).46 Personnel density requirements demand a ration of one pre-school teacher per 14-18 children in Norway, whereas in Sweden the corresponding figure was one per 5. children. The sizes of children’s groups remain unregulated in Norway, whereas they are delimited to 15 in Sweden. Similarly, requirements for professional qualification were less strict in Norway. Due to larger groups of children per trained personnel, the density of required qualified personnel was lower in Norway than in Sweden. Educated personnel have been required in Sweden since 1957 and in Norway since 1975.47

65 The 2005 day-care settlement The relative independence of associational providers from operational control changed with the 2005 “day-care settlement” (barnehageforliket). The day-care settlement implied that associations had to accept municipally determined operation requirements. Associations became obliged to cooperate with municipalities concerning the length of opening hours, the age groups of children attending, and the placing of children according to municipal waiting lists. Municipalities, however, were encouraged to consider day-care providers’ “variety and particularity” (Day-Care Act 2005). The day-care settlement also implied that parent fees were to be limited to maximum levels in 2005. It was feared that the introduction of maximum levels for parent fees would eventually lead to financial problems for associational providers. Until 2005, associations had no right to municipal subsidies and, as a result, demanded higher parent fees than municipal providers (Innst.S.nr. 250 2002-2003, Ot.prp.nr. 68 1993-94, cited in Blom-Hansen 1998). In 2002, for example, parent fees for an average full-time place in associational day-care amounted to 3400 NOK in comparison to 2700 NOK in municipal day-care, which corresponded to 125% of municipal day-care fees (Innst.S.nr. 250 2002-2003). As a recent study shows, associational providers managed to reduce their fees to the legal maximum levels posted in 2005. However, unlike many municipal providers, associations usually did not provide for income moderations for parent payments (Rauan 2005).48 At the same time as associations were subject to increased demands coming with the day-care settlement, they also gained access to municipal day-care subsidies. Associations had often received additional municipal funding before 2005, but a considerable number of associations were subsidised only through statutory grants. As many as 30 to 50% of all associations operated without any further municipal subsidies (see Appendix A table A14, Blom-Hansen 1998, Nicolaysen 2001, NSD). This situation was changed by the central state in 2005 by explicitly formulating a right of associations to municipal subsidies, thus providing associations with similar rights to municipal funding as public providers. Although the conditions for associational provision changed considerably, a decrease in associational day-care participation was maintained unabated after 2005.

66 So far it has not been mentioned that conditions for public subsidies differ between associational and municipal providers in both countries. Municipalities subtract an “administration tax,” which leaves associational providers with on average 80% of the subsidies municipal providers receive in Sweden, and 85% in Norway (Skolverket 2005:261, Skolverket 2003:231, KD Forskrift om likeverdig behandling av barnehager i forhold til offentlige tilskudd 19.mars 2004, Day-Care Act 2005). Whether these subsidy levels imply an unjustified diminishing of public subsidies for associations is debatable. Associations usually regard 80-85% to be as unjustifiable, whereas public authorities claim the difference to be well-founded due to the administrative tasks conducted by municipalities such as planning and coordinating day- care provision for all providers within the municipal area. Half of all Swedish municipalities also administer the collection of parents’ fees on behalf of associations or provide for advanced training courses (Strandbrink and Pestoff 2006, Skolverket 2005:261). The institutional changes of associational day-care provision in Sweden and Norway are summarised in table 4. • Table 4: Statutory funding and regulation of day-care

1940- 1960- 1970- Sweden Day-care Profession 1963 and 1966 1973 regulations al leading Subvention for half- and Municipal day-care plans and personnel fulltime day-care increased; obligations to provide day-care Subvention for short-time day- for six year old children care abolished Regulations 1944 1973 concerning Eligibility Associations not regarded as associations to statutory equivalents to municipal day- subsidies care; Subsidies for municipal short- time day-care facilities only Further School age raised to 7 regulations 1940- 1960- 1970- Norway Day-care Day Care Act 1975 Regulations Professional leading personnel Regulations 1963 concerning Eligibility to statutory subsidies associations

Table 4 (continued): Statutory funding and regulation of day-care

1980- 1990- 2000- Sweden Day-care 1991 regulations Municipal obligation to provide day care Regulations 1983 1994 2003 concerning Delimitation of eligible Equal treatment of public Maximum limits for associations associations and non-public providers; parent fees 1987 Eligibility to municipal Subsidies for municipal subsidies short-time day-care facilities only 1980- 1990- 2000- Norway Day-care 1982-97 2003-5 Regulations Subventions for half-/ full-time Maximum limits for and for younger children increased parent fees; Extra subsidies for conversions 1991-97 Municipal day care plans Regulations 1982 2005 concerning Widening of control Eligibility to statutory associations rights: personnel and municipal selection and curriculum subsidies determination Further School age lowered to 6 regulations from 1991 Kontantstötte 1998

Summary Especially during the 1970s and 1980s, statutory day-care subvention and regulation can be regarded as having been more supportive of associational day-care provision in Norway than in Sweden. In Norway, associations had uninterrupted statutory funding with the same conditions as municipal providers while, in Sweden, their eligibility was restricted during the 1970s and 80s, two decades of overall strong day-care expansion. Thus, associations could develop to similar conditions as municipal providers in Norway, whereas associations had less support than municipal providers in Sweden. Likewise, funding for short-time day-care had been more generous in Norway than in Sweden, which can be regarded as supportive for associational day-care provision, due to associations’ predominant provision of short-time day-care. Furthermore, Norwegian associations were more independent than their Swedish counterparts. Whereas Norwegian associations were directly entitled to statutory funding if they met general quality standards, Swedish associations had the additional demand of having 68 to fit into municipal development plans for day-care. Furthermore, Norwegian religious associations were explicitly excluded from certain operational standards. They were allowed to both educate children according to their worldviews (irrespective of parents’ wishes and legal prescriptions about curricula standards) and to disregard labour laws when choosing personnel according to their convictions. In addition, Swedish regulations were more demanding than Norwegian regulations, for example, requiring higher professional standards of day-care staff or longer opening times. Since the 1990s, however, statutory regulations converged between the two countries. In Sweden, associations’ eligibility became equal with that of public providers for all forms of statutory subsidies in 1991 and, in 1994, they became eligible for municipal subsidies. In Norway, associations remained equal with municipal providers and likewise became entitled to municipal subsidies in 2005. Parallel public regulation of associational provision converged by remaining high in Sweden and by enhancing municipal control over associational provision in 2005 in Norway. I have argued that the development of associational day-care • provision was strongly influenced by statutory policies. Parallel to the development of statutory regulations, the level of associational day- care activity differed between the two countries until the 1980s, but has since then converged. Among the regulative settings, eligibility for statutory funding may have been most important for the overall level of associational activity.49 It is generally thought that the growth of associations in Norway since the 1970s and in Sweden since the 1980s was enabled mainly by statutory subsidies. Relatively more generous funding for short-time day-care and more equal conditions for associational providers in comparison with municipal providers in Norway than in Sweden may have contributed to a higher level of associational activity in Norway until 1990. As to operational regulations, it could be argued that they do not seem to have been principle obstacles for the overall level of associational day-care activity. In both countries, operational control increased over the decades without diminishing the level of associational activity. However, operational control can have implications for some associations’ interest in becoming engaged in day-care and, which can, in turn, have implications for the composition of associational providers. Eventually, operational

69 prescriptions can conflict with associations’ own interests in day-care provision, and, consequently, become disincentives for some associations. A relatively greater independence of associations to operate their facilities in Norway could have provided them with the necessary elbow room to provide day-care along particular interests. However, the development of associational day-care cannot merely be related to access to statutory subsidies. Funding regulations tended to be more favourable in Norway, but Swedish associations were not excluded over the entire period. Like their Norwegian counterparts, Swedish associations were eligible for statutory funding during the 1960s and since the 1990s, but Swedish associations made fewer efforts than their Norwegian counterparts: they put a great deal of effort into day-care provision. The next section will attempt to estimate associations’ interests in providing day-care by describing the different engaged associations in day-care provision in Norway and Sweden.

4. Associations in day-care As discussed above, associations’ efforts to provide day-care differed between Norway and Sweden until the 1970s and, in addition, in Sweden before and after the 1970s. In the following, it will be investigated whether such differences in associational engagement might be based on associations’ interests in providing day-care or on more or less supportive statutory regulations for associations’ day-care interests. It could be assumed that associational founders with strong interests that differ from ‘mainstream’ day-care are more eager and, therefore more likely, to insist on maintaining their own facilities than founders, which regard their day-care interests sufficiently covered by other providers.50 The strong day-care interests of associations could be based on, for example, interests in a specific function of day-care, interests, in the proliferation of a certain worldview or pedagogy or simply an interest in the maintenance of day-care as a main field of their organisational operations. However, one has to keep in mind that associations’ efforts may not only depend on their interests, but also on their organisational capacity to provide day-care. If associations lack committed people or knowledge to found day-care facilities, they may simply fail to realise their day-care interests.

70 In order to achieve a first impression of Swedish and Norwegian associations’ interests in day-care, associations’ operational parameters and associations’ own interest announcements are used. In addition, officially reported categories of associations are taken as approximate information over associations’ day-care interests (a more in-depth description of some associations’ interests will be given in section 6). The efforts of associations to provide day-care with specific provision parameters are assumed to be indications of underlying functional interests of their owners. Variations of operational aspects, such as the length of opening hours or the attending children’s age groups, are considered to be indications of varying interests of owners in related functional outcomes. Short opening hours are interpreted as indications of an interest in an educational or pedagogical functioning, whereas provision for specific day-care clientele is regarded as associations’ interests in catering for respective social segments, be they socio-economic or ethnic segments. Additionally, categories used by official statistics are taken as sources of information. • Official categories are regarded to inform indirectly over associations’ day-care interests, because they often provide information over either ownership relations of welfare associations to umbrella organisations or day-care associations’ juridical organisation forms. Umbrella organisations, such as the Red Cross, Fröbel Association and the Norwegian State Church, usually articulate, for example, broader philanthropical interests, pedagogical day-care interests or broader denominational interests. Such interests and intentions are assumed to influence member associations and are considered to be approximations of the main thrusts of their interests in day-care. Day- care facilities owned by the Fröbel Association, for example, are assumed to pursue a pedagogically oriented provision of day-care. Day-care facilities owned by the Red Cross are assumed to follow general philanthropical interests in the shaping the operation of their day-care facilities. Official categories further classify associations according to their juridical organisation forms, such as cooperative, firm, ideational organisation, et cetera. Juridical organisation forms are more difficult to interpret in relation to the interests of the owners of an associational facility, because owners may simply choose one or the other juridical

71 form due to related tax regulations or subsidy regulations. However, juridical forms provide some indication of the owners’ interests. The organisation of a day-care facility as an ideational organisation may indicate that the interests of the owners lie in providing day-care with a certain curriculum, rather than the provision of day-care in order to facilitate pure work-related supervision. In contrast, the juridical form of a parent cooperative provides information over the composition of owners and the specific relation between owners and clients, which may indicate a high interest of the parent owners in controlling day- care for their children. Based on these deliberations, the question of how interests may have influenced the diverging efforts of associations in Sweden and Norway will be discussed throughout the main part of this section. Afterwards, statutory regulations of associational welfare will be discussed in terms of how supportive they were to the specific day- care operations and interests of main associational providers.

Phase I Sweden until the 1970s In Sweden, associations had been eligible for statutory subsidies from 1944, but they almost ceased to operate until the 1970s. Their share of day-care provision collapsed from 94% in 1943 to 2% in 1970. What kinds of organisations were responsible for associational day-care during these early decades and which interests may have motivated their engagement and later withdrawal? Associational facilities in the 1940s belonged to a large variety of owners, including philanthropic associations, such as the Swedish Red Cross, the Salvation Army, housing cooperatives, unspecified owners, missions, churches, and the Swedish Housewives’ Federation; and pedagogical associations such as the Frederika-Bremer Association, Fröbelförbundet and others (Svenska Fattigvårds- och Barnavårdsförbundet; see also Appendix A table A16). Associations that may have had particular interests in day-care, such as pedagogical associations or churches, were not predominant among associational providers.51 Most salient were associations belonging to the Red Cross, the Salvation Army or housing cooperatives, which had their main activities and assumed interests in other areas than day-care. However, none of the active associations owned more than a few facilities.

72 Many of these associations were engaged in short-time day-care as described in the foregoing section. Religious- and pedagogic-oriented associations can be assumed to have deliberately chosen this form of day-care provision. However, it remains an open question if the predominantly active philanthropical and housing associations provided short-time day-care out of their interests or because they had insufficient organisational capacities. Therefore, the picture of associational day-care in Sweden is of a large variety of active associations and related interests. A number of the active associations could be assumed to have a strong interest in maintaining their activity alongside mainstream day-care in order to provide pedagogical or denominational short-time day-care, whereas philanthropic or housing associations may not necessarily have had a special interest in maintaining day-care provision as one of their activities. However, of the once active associations that owned 324 day-care facilities in 1943, only a few remained (113) in 1970. This also seems to indicate a broader lack of interest on the behalf of associations to maintain their activity or, alternatively, indicates the insufficiency of their organisational capacities. • Further indications of a lack of associational interest in preserving their day-care engagement alongside public providers, are associations’ comments to Swedish government official inquiries.52 Surprisingly, even though associations were, at this point, dominant in day-care, only a few associations publicly commented on the plans to introduce day-care subventions in 1944. It might be expected that associations would have seized the opportunity to publicly highlight their engagement, yet in their submitted comments, they responded positively towards stronger public engagement. The Red Cross, Fröbelförbundet, YWCA Pedagogical Institute, Sveriges Husmodersföreningars Riksförbund, and the national umbrella organisation Swedish Poor and Child Care Union all welcomed the prospect of the state taking over responsibility for day-care provision. The Red Cross explicitly approved stronger public surveillance of associations’ day-care facilities, which, according to the Red Cross, had too often been of poor quality (Prop. 1943:339). Furthermore, none of the commenting organisations expressed a desire for associational provision in particular or emphasised advantages with their own activity. However, in 1973, when day-care for six year old children was extended and associations were not regarded as an eligible provider

73 category for them, the Swedish State Church alone protested against the proposal, that its children related activities were not regarded as an equivalent to public provision (Prop. 1973:136). A further indication of a less developed interest to preserve associational activity and approval for public provision is that some of the associations even applied to become municipally organised.53 However, another explanation could be that this was due to insufficient capacities. In sum, Swedish associations made no specific efforts to enlarge their engagement with the help of statutory subsidies, nor did they express a strong commitment towards particular associational solutions for day-care. The reluctance of associations to prolong or extend their day-care engagement after increased public involvement could partly be interpreted as a lack of interest among larger philanthropic organisations to maintain day-care provision (e.g., the Red Cross); they were more active in welfare areas other than day- care. In contrast, associations belonging to the Fröbel Union, which undoubtedly held a specific interest in day-care provision, may either have regarded their interests covered by public provision (which indeed came with high pedagogical quality standards) or may simply have lacked the necessary organisational capacity to uphold their activity (eventually due to a shift of statutory subvention from short- time day-care to half- and fulltime day care).

Norway until the 1990s In contrast to Sweden, Norwegian associations steeply increased their facilities after the enhancement of statutory funding in 1975. Associations were able to increase the number of their facilities from 207 in 1973 to 1721 in 1990 and to maintain a large share of day-care provision despite the rapid overall expansion (it decreased from 49% in 1973 to 39% in 1990). Certain categories of associations showed especially strong engagement. More than 60% of associational providers and about 30% of all day-care providers were private organisations or parishes of the Norwegian State Church. They dominated associational day-care provision in Norway54 until the beginning of the 1990s (see figure 6 and Appendix A table A19). As Lea has shown, the category private organisations was comprised mainly of conservative and religious organisations such as the Norwegian Housewives’ Federation and denominational organisations in the 1970s and 1980s (Lea 1982, also

74 St.meld.nr 89 1961-62). Obviously, these organisations had a strong interest in day-care provision. The efforts of the Housewives’ Federation, for example, can be seen as an extension of one of their main organisational interests. The Housewives had advocated pedagogical day-care expansion in Norway since the 1920s and likewise had a history of running day-care facilities (Solbraa-Bay 1970, 1985). Their engagement can be regarded as being based on an interest in maintaining and extending a main area of their activities (see also section 6). Denominational organisations also expressed strong motivation to provide day-care. Their stated motivation was to establish a particular associational day-care, based on religious interests, as expressed in internal circulars published in response to the introduction of state grants: “The time has come to focus on pre-school children in our parishes. […] Seize the chance! Within a few years, public day-care will be reality in our country also, and then we as parishes will have less opportunity to influence it” (Norwegian Sunday School Union). The Norwegian Lutheran Mission informed their members: “[…]this means that we now have a golden opportunity to get Christian day- 55 • care; within a few years it could be too late” (see also section 6). Thus, Norwegian associations showed a strong commitment to provide day-care with a pedagogical or Christian orientation in combination with an interest to remain active as associations along with other providers. This commitment contrasts with Swedish associations’ indifference in maintaining their activity and may have been one reason for the strong increase of associational day-care activity in Norway.

Phase II Since the 1980s, associational day-care provision converged in Sweden and Norway. The convergence concerned the share of associations among day-care providers and the composition of associations. Today, Sweden has about half as large a share of associational day-care provision as Norway and, in both countries, various associations are active and no type of association is as predominant as religious-conservative associations had been in Norway before the 1990s.

75 Sweden since the 1980s The efforts of associations changed in the 1980s and 1990s. Since the 1980s, there has been a slow and constant rise in associations’ share of day-care provision. It is particular to Swedish development since the 1980s that associations engage in day-care alongside public providers. This reverses the development of the 1940s to 1970s, when associations’ day-care activity dissolved. Day-care facilities are organised by a large variety of actors including parent cooperatives, employers, the Swedish State Church, Waldorf associations, and employer cooperatives. The composition of associations reveals no single type of association that could be regarded as predominant (see figure 5 and Appendix A table A17). A wide variety of interests could be derived from the categories provided. Employers, for example, can be assumed to provide day- care in order to attract personnel, employee cooperatives to have an interest in self-employment, the Swedish (State) Church to have an interest in educating children in Christian beliefs, and the Montessori and Waldorf associations, which are linked to pedagogical movements, to have an interest in a specific pedagogical orientation of day-care. The most salient type among non-public providers is parent cooperatives. A case study of the Education Department indicates that parents’ interest in organising into parent cooperatives is a specific interest in self-operated day-care (Sundell and Ståhle 1996). According to this study, parents said that the organisation of day-care in the form of a parent cooperative was the most important reason for choosing such facilities above others. This distinguishes them from all other parents questioned in this study, who said that they made their choices out of practical and quality considerations (Sundell and Ståhle 1996).56 The study further indicates that the reason for the relative success of parent cooperatives seems to be the high level of engagement by their owners. This conclusion is also corroborated by other studies (Strandbrink and Pestoff 2006, Skolverket 2005:261). In sum, Swedish developments since the 1980s seem to be influenced by various smaller associations, which all seem to have specific interests that motivate them to become involved in day-care.

100

20 share of day-care by facilities provider, % 0 1943 1960 1970 1980 1990 1995 2000

public sum churches, housing cooperatives, employers parent cooperative employee cooperative • other private private for-profit

Figure 5: Sweden: Day-care facilities by provider, shares in % Source: see Appendix A tables A16 and A17

Norway since the 1990s In Norway, associational provision remained at a high level, but the composition of associations changed in the 1990s. Private organisations, parishes and denominational associations, which had so far been predominant among associational day-care providers, saw their shares fall from around 30% until 1990 to below 10% during the 1990s (see figure 6 and Appendix A table A19 and note 57). These associations became outnumbered by a rise of other associational types during the 1990s. The two largest categories were parent associations and individuals. Parent cooperatives and individuals account for 14% and 15% respectively of all day-care facilities, and each accounts for about one-third of associational facilities.57

10 share day-care of by provider, facilities %

0 1975 1980 1984 1988 1990 1996 2000

public employer parish, private organisation parents, individuals miscellaneous

Figure 6: Norway: Day-care facilities by provider, shares in % Source: see Appendix A tables A19

Reasons for this change of associational provider composition could have been the divergence of functional interests, especially interest in providing day-care for younger children and with longer opening hours. Since the middle of the 1980, most providers prolonged their opening hours and increased their provision for younger children, whereas private organisations, parishes and denominational associations continued to provide foremost short-time day-care catering for older children (see table 5).58 The maintenance of a distinct form of day-care by these organisations could be regarded to be based on their interests in providing day-care oriented to a conservative, religiously inspired family model. In contrast, parent cooperatives and individuals seem to have been especially eager to provide full-time day-care for younger children: their efforts resulted in an above-average provision of such

78 day-care and, furthermore, exceeded the efforts of municipalities in this respect. In sum, the altered composition of associational day-care providers in Norway could be interpreted to be based on different interests in day-care provision. Former predominant associations belonging to the Housewives’ Federation or to churches and lay organisations were more interested in day-care with shorter opening hours and for older children. This indicates that these associations had an interest in providing day-care as a supplement to traditional family care rather than providing day-care to support female labour participation (see also section 6). Newer types of associations in Norway provided day- care with long opening hours and for younger children, offering child supervision for working mothers.

•

79 Table 5: Norway: Younger children with fulltime day-care in facilities by ownership, shares of enrolled children, %

Public Associations All Enrolled organisations children, total Municipalities, Private organisations2 Parishes, Employers Parents Individuals counties, state denominational organisations2 1980 9 2 1 11 1 na 6 78181

80 1986 17 8 5 28 6 Na 13 107790

1990 22 13 7 32 37 26 20 139350

1995 26 8 10 43 28 41 27 188213

2000 26 16 13 41 29 35 28 189837

1 younger children 0-3 years old, fulltime: 40+ hours per week 2 categories changed over time (see Appendix A table A20) Source: NSD

Summary The composition of associations, based on official categories, differed between the first and second phase in Sweden and Norway and coincided with variations of associations’ operational parameters. In combination with associations’ own interest announcements, these changes are interpreted as indications of alternating interests in day- care provision by the main active associations, resulting in different efforts by these associations. During the first phase of day-care expansion in Norway, religious- conservative day-care associations were heavily engaged in short-time day-care provision for older children. According to their own announcements, their main interest was to provide day-care based on Christian values as an associational alternative alongside public day- care. In Sweden, associations also provided short-time day-care, but they seemed to lack the same dedication to providing alternatives to publicly provided day-care. Philanthropic associations’ minor eagerness could be regarded to rest in their otherwise well-developed • welfare activities, which they may have prioritised above engagement in day-care, but it remains unclear if pedagogical or denominational associations were interested in prolonging their engagement or if they were more supportive of public provision than their Norwegian counterparts. During the second phase, parent cooperatives became prevalent in both countries, alongside other categories of associations. Parental associations and other involved associations currently provide day- care with longer opening hours and for younger children, indicating efforts to provide day-care services for working mothers. Some Swedish studies provide further indications of interest in parent cooperatives. According to case studies, parents are interested in parent cooperatives due to a wish to own and operate day-care organisations for their own children, which seems to indicate an interest in controlling their children’s upbringing. According to a national parent survey, parents of children attending non-municipal day-care more often have an academic background and are more often of Swedish origin, indicating that non-municipal day-care facilities cater to families with specific social backgrounds. How supportive have the main regulative settings been for the specific interests of associations in Norway and Sweden?

81 In Norway during the first phase, conservative and religious associations eagerly engaged in day-care after institutionalisation of statutory transfers to associational providers. They mainly provided short-time day-care for younger children. The regulative setting has been especially supportive of the most active types of associations. First, it was favourable for associational self-regulation, allowing associations to hire personnel and determine curricula according their convictions. Second, statutory funding for short-time day-care and care for older children was more generous, possibly reflecting the interests of religious and conservative associations, which were primarily engaged in this type of care. During the second phase, the composition of associational providers changed in Norway. Today, parent cooperatives are the most salient type, while religious-conservative-oriented associations are no longer dominant. Religious-conservative providers may have been reluctant or unable to provide full-time day-care and day-care for younger children, while parent cooperatives seem to have been eager to provide such care. Today, associational providers even outperform public providers with respect to length of opening hours and care for younger children. The altered pattern of prevalent types of associations may reflect, first, an adjustment of statutory subsidies from the 1980s until the middle of the 1990s, which improved funding for day-care for young children and longer opening hours and, second, provided extra funding to support transformation of short-time day- care for older children to long-time day-care for younger children. In Sweden, associations did not make great efforts to engage in day-care during the first phase, although statutory funding was provided until the middle of the 1970s. The few engaged associations predominantly provided short-time day-care for older children, but even these associations further reduced their activities during the 1960s. The further decrease of associational provision during the 1960s may reflect a reduction in public funding, which was removed for short-time day-care facilities, but it could also be an outcome of associations’ acceptance of public provision. Since the second phase after 1980, however, associations started to grow, which can be regarded as an outcome of both increased associational engagement and stepwise enhanced access to statutory funding. Reasons for greater engagement by associations could be that, in contrast with the first phase, there are more associations involved that

82 have day-care as their main activity. Parent cooperatives, for example, can be regarded as associations of parents who have a specific interest in controlling day-care provision for their children themselves. In addition, increased associational efforts may also reflect a decreased acceptance of public provision, which, after decades of predominance in the day-care market, may have increasingly revealed short-comings from the perspective of some day-care interested individuals and organisations. The most important regulative improvements that have promoted an increase in associational day-care during the 1980s and 1990s in Sweden were, arguably, the broadening of the range of eligible categories of associations, and increased equality between associational and public providers with respect to short-time subsidies, personnel subsidies and municipal subsidies.59 So far the focus had been on associations’ interest in being active in day-care and on the regulative settings for their activities. However, one has to keep in mind that the relative prevalence of associations depends not only on these two factors, but also on factors such as associations’ organisational capabilities and, not least, on parental • interests in how and through whom day-care is supplied. The relative success of the associations investigated in this study may have been based on their ability to organise day-care provision more efficiently or effectively than competing associations. It remains unclear, for example, whether voluntary work contributed to differences in the development of associational providers. The relatively low levels of activity among churches and pedagogical associations in Sweden, for example, could have been due to their lower attractiveness to voluntary workers compared with their Norwegian counterparts. Likewise, the decrease in religious associations and increase in parent cooperatives in Norway could be due to an eventual decrease in the ability of former religious associational providers to attract volunteers and an increase in voluntary work for parent cooperatives. Evidence of the scope of voluntary work is scarce, but newer case studies indicate that parents do engage in voluntary work; it can take the form, for example, of one week per year of cooking or cleaning duties in Sweden, and organising day-care excursions in Norway (Walter and Wächter 2000, Strandbrink and Pestoff 2006, Skolverket 2005:261, Forbrukerombudet 2005). It could be speculated, that parents increasingly preferred to volunteer for their own organisation

83 than for a religious organisation, thereby increasing the competitiveness of parent cooperatives over denominational associations. Furthermore, assistance from other organisations may have supported the growth of parent cooperatives. Swedish parent cooperatives, for example, were assisted by the Swedish national cooperative union (KF), two national housing cooperative organisations (HSB, Riksbyggen) and a large national insurance cooperative (Folksam), which initiated a service organisation for parent cooperatives in 1987. This service organisation, “Our day-care centre” (Vårat dagis), provided legal and administrative help for starting parent cooperatives (it was abolished in 1997, NOU 1988:30). Parental demand for certain forms of day-care or demands for certain types of associations are other factors influencing the success of associations. In Norway, increased levels of female employment may have led to an increased demand for day-care during work hours. Such shifts in demand, as expressed by political pressure, parents’ own efforts, and the setting of market incentives, may have put associations providing short-time day-care at a disadvantage compared to those providing day-care during work hours. Parental demand for non-municipal day-care may also have changed the opportunities for associational day-care providers in Sweden. A national parent survey shows that 30% of parents with children in non-municipal day-care would prefer municipal day-care for their children, and that 30% of parents with children in municipal day-care would prefer non-municipal day-care (Skolverket 2000). Taken together, this points, indeed, to a change in Swedish parents’ attitudes towards public day-care. These surveys show further that parents of children in non- municipal day-care are more often of Swedish origin and have higher educational levels. Only 3% of children of parents born abroad attend non-municipal day-care compared to 10% of children of Swedish parents. Further, only 3% of children with non-academic parents attend non-municipal day-care compared to 18% of children with academic parents. These patterns could be interpreted as a demand among well-educated Swedish parents for socially distinguished day- care (Skolverket 2000). In sum, various factors have influenced the development of associational participation in day-care provision in Sweden and Norway. Of these factors, associations’ interests in becoming engaged

84 and regulative settings for associational participation have been described here in greater detail. According to the hypothesis of this thesis, the forming of regulative conditions for associational day-care, which will more or less favour some associations’ interests in day-care provision, is an object of partisan interest. Political parties that share the interests of specific associations will attempt to mould the formulation of regulative settings for associational welfare participation in order to favour day- care provision by those associations, while political parties that do not share the interests of associations will oppose such attempts. In order to investigate the policies of different parties, I will focus in the next section on how partisan policies have influenced the formulation of the above-described public conditions for associational day-care provision. Have parties had conflicts over the regulation of associational day-care? Which parties were the main opponents or advocates of increases or decreases in associational subsidies and operational independence?

• 5. Partisan policies In general, political parties in Sweden and Norway agreed upon state intervention into day-care.60 The first central state day-care policies in the 1940s in Sweden and the 1960s in Norway were supported by both governing and opposing parties. Afterwards governments of alternating political colours let day-care subsidies grow, especially since the 1960s and 70s. Partisan support for day-care subsidies was triggered by interests in female labour market participation during times of economic growth and tight labour markets, when major economic interest organisations also agreed upon national subsidies for day-care (see Appendix B). When the economic situation changed, with rising unemployment, slowed growth rates and budget deficits in the 1980s, political support for day-care subsidies nevertheless continued across the right-left spectrum. This reflected strong popular support for day-care expansion at a time when female labour market participation had risen considerably in both countries. Since the 1970s, national policy goals for day-care became ambitious in both countries, expressing a commitment to day-care provision according to demand. In Sweden, the aim has been that “the general goal for expansion shall be to […] provide a place in […] day-

85 care for all children of pre-school age of employed or studying parents […]” (Prop. 1975/76: 92) and, in Norway, “that all children, whose parents wish it, should have a day-care place” (Ot.prp.nr. 23 1974-75, cited in Blom-Hansen 1998). However, while they agreed on the basic issue of day-care expansion, political parties disagreed about specifications for day-care policy. Controversies arose, for example, about the extent to which day-care policies should be formulated to substitute for traditional family care; right-wing governments tended to support policies for short-time day-care and home care allowances, while left-wing governments were more supportive of day-care for children under three years old and full-time day-care (Hinnfors 1992, Ot.prp.nr. 75 1981-82, St.prp.nr. 53 1997-98). Furthermore, partisan controversies arose about how standardised day-care should be provided for. Left- wing parties, for example, introduced limits on parent fees for day- care after 2000, while right-wing parties opposed such limits. In Norway, a further controversy arose about educational standards, where a right-wing government introduced Christian education for all forms of day-care, which was opposed by left-wing parties (Ot.prp.nr. 75 1981-82). Besides political controversies about general day-care policy, there were also disagreements about associational day-care provision, which is the main area of interest here. In many respects, the above- mentioned controversies were connected to associations’ day-care provision, but additional arguments arose about subvention and regulation of specifically associational provision, to which I will now turn for a description.

Sweden National intervention in day-care provision started in 1943, when national government subsidies for day-care were introduced. The government draft included associational day-care provision along with municipal day-care, that is, associations became eligible for statutory subsidies under the same conditions as municipalities. All parliamentary parties supported the introduction, and the government draft for day-care expansion was approved without significant controversy (Prop. 1943:339).61 During the 1960ies, partisan controversy arose about the reduction and later abolition of statutory subvention for short-time day-care

86 facilities, which may have contributed to the decrease of associational provision during the 1960s and 1970s. These changes were introduced by social democratic governments. By contrast, right-wing parties, such as the Conservative Party, the Liberal Party and the Centre Party, voted against the proposals. The argumentation of right-wing parties was that short-time day-care was a pedagogically valuable form of day-care and should continue to be eligible for statutory funding (Prop. 1966:54, Prot. II 1966:21, Mot. II 1966: 817, Mot. I 1966: 665,666). At the beginning of the 1970s, further partisan controversy arose around the introduction of municipal obligations to develop day-care plans and to provide day-care for six year old children. The social democratic government did not regard associational day-care as an equivalent to municipal day-care and, thus, proposed that municipalities exclude associational day-care activities from municipal planning in fulfilling their obligations. By contrast, right- wing parties claimed that municipalities should be allowed to incorporate associational activities in order to fulfil their obligations and voted against the proposal of the social democratic government • (SFS 1973:1205, 1207, Prop. 1973:136, SoU 1973:30). These parties mentioned especially associational day-care activities of the Swedish State Church, other denominations and ideational organisations, which should be considered as a part of municipal day-care obligations. Furthermore, the social democratic party re-introduced funding for short-time day-care, but only for municipally organised short-time day-care. Right-wing parties, on the other hand demanded parity also in this respect for associational provision. In the following years, right-wing parties exerted further pressure against this regulation, claiming that associational provision was a culturally and pedagogically valuable form of day-care provision, not least due to associations engagement in short-time day-care (Mot. 1975: 342, 351, 477, SoU 1975/76:28). In 1976, the social democratic government partly changed its position. It allowed municipalities to decide upon an incorporation of associational provision as long as associations fitted into municipal day-care plans. However, the social democratic government abolished funding for short-time day-care again (SoU 1975/76:28). The Conservative Party opposed this proposition and claimed instead direct eligibility for associations to statutory funding and a legally

87 sanctioned right of associations to be incorporated (SoU 1975/6:28; Prot. 1975/6:95). During the 1980s, associational eligibility remained a subject for partisan controversies. From the beginning of the 1980s, social democratic governments started to issue “positive lists” of associations, which were eligible to statutory funding. In 1983 and 1984, eligible associations had to be either organisations grounded on ideational notions (ideell förening) or parent cooperatives. Furthermore, the Social Democratic government introduced a special personnel grant, which was available only to municipal providers and delimited associations access to short-time subsidies. In ongoing debates around associational provision during the 1980s, right-wing parties pressed for an equalisation of associations with municipalities concerning personnel grants and short-time subsidies. Furthermore, right-wing parties advocated an extension of the circle of eligible associations, such as the Swedish State Church, Christian groups, Montessori associations and Waldorf associations (SoU 1983/4:12, 1985/6:35, Mot. 1986/7: 605, 608, 639; Prot. 1982/83:126, Prot. 1983/84:41). Their argumentation for a liberalisation was that private day-care would allow consumers a choice of pedagogical and educational variation and eventually lead to reduced public costs for day-care. It was seen as a “parent’s right to choose alternative day-care that addresses special needs or has a particular pedagogical or ideational orientation” (Conservatives) and it was argued that associations “use experiences from alternative pedagogies” (Liberal Party). In addition, it was stated that “voluntary energy should be used,” which “would provide room for great variation and creativity” and that “several thousand hours” of voluntary work could be drawn upon “to get [care] cheaper carried out” (Centre Party) (Prot. 1982/83:126; Prot. 1983/84:41).62 By contrast, the Social Democratic Minister of Social Welfare defended the change, stating that he “suspected that the reason for the Conservatives’ privatisation campaign [was] that in a tense economic situation they want[ed] to take money away from municipal day-care and shuffle it over to private day-care for the highly-paid.” However, he opened up the possibility for associational providers by arguing that the government wanted “to prove new forms to improve parental teamwork, support parent cooperatives and various associational efforts. Provision should be encouraged if it meets demands of equity and good quality” (Prot. 1983/84:41).63

88 The governing Social Democratic Party conceded to some demands, for example, abolishing the special personnel grant for municipalities and including personnel cooperatives as eligible associations in 1990 and the Swedish State Church in 1990 and 1991 (SFS 1987:860, SFS 1990:1188, SFS 1991:237). When the right-wing coalition returned to power in 1991, they revised the changes made by the Social Democratic Party and made all types of non-public providers eligible for statutory subsidies again. Furthermore, they equalised non-public providers with respect to eligibility for municipal subsidies and all forms of statutory subsidies. Subsequent social democratic governments made no lasting changes concerning associational providers’ rights towards public subsidies. In 2002, however, the social democratic government anew introduced a controversial regulation. Day-care providers became obliged to keep their demands for parents’ fees under specific maximum levels (maxtaxa). The introduction of maximum levels, which could have resulted in economical problems for some associations, were opposed by the parties of the right-wing opposition (Prop. 1999/2000:129).64 • In sum, political parties disagreed over several regulations concerning associational provision see table 6 for summary of partisan policies). The Social Democratic Party favoured a reduction of subvention for short-time facilities and preclusion from these for associational short-time facilities, while right-wing parties opposed against these regulations. Furthermore, the Social Democratic Party precluded associations from being part of obligatory day-care plans, whereas right-wing parties demanded an inclusion. Social democratic governments also precluded a number of associations from statutory subsidies by issuing “positive lists” of eligible associations. By contrast, right-wing parties demanded and later obtained the inclusion of all providers as a means to as a means to enhance educational and pedagogical variation and to reduce public expenses for day-care via expected increases in voluntary work. Thus, the overall impression is that social democratic governments tended to hold a distanced position towards associational day-care provision. The Social Democratic Party tended to pursue policies, which either delimited associational eligibility or put associational participation at risk of losing public resources to fund their activities. By contrast, right-wing parties defended associational participation

89 and pursued policies, which assured and widened associational access to public funding.

Table 6: Conflicts around associational day-care provision, Sweden

Day-care laws Policy items Supporting Rejecting and parties parties administrative laws 1943 Introduction of subsidies for All parties associations 1963 Reduction of subsidies for short-time SAP M, FP, C funding 1966 Abolition of subsidies for short-time SAP M, FP, C facilities 1973 Municipal obligation to develop day- SAP M, FP, C care plans without admission of associational day-care 1983 Reduced eligibility for associations SAP M, FP, C

1993 Full eligibility for associations M, FP, C, KD SAP

2002 Maximum levels for parent fees SAP M, FP, C, KD

Social Democratic Party (SAP), Conservative Party (M), Liberal Party (FP), Centre Party (C), Christian Democratic Party (KD)

Norway Over the whole period of statutory day-care intervention, associations’ eligibility for public subsidies was in principle agreed upon across party lines. Any political conflicts that arose about associational day- care provision concerned the types of provision that should be eligible and the degree of independence associations should have in operating their day-care facilities.

1963-1975 The introduction of central state subsidies for day-care in 1963 and subsequent extensions until 1975 were proposed and introduced under labour governments and agreed upon across party lines.65 The Day-Care Act of 1975, which marked the shift to greater public responsibility for day-care expansion, was likewise supported by all

90 parties. The consensus on enhanced statutory intervention included associational eligibility for statutory subsidies: every party accepted in principle the eligibility of associations for statutory subsidies along with public providers (Ot.prp.nr. 23 1974-75, Innst.O.nr. 69 1974- 75).66 Nevertheless, indications of potential conflict over associational day-care provision can be found in the reservations expressed by the Christian People’s Party concerning the formulation of general humanist standards as main guiding principles for the operation of day-care in 1975. Instead of the stipulated guiding principles, which included rather general humanist formulations and which were supported by all other parties, the Christian People’s Party demanded a formulation emphasising Christian values and norms, indicating their interest for shaping day-care provision towards specific normative stances (Innst.O.nr. 69 1974-75, Ot.prp.nr. 23 1974-75). Further indications of potential conflict were divergent arguments for statutory intervention in day-care. On the one side, the Conservative Party and left-wing parties most clearly argued for day- care expansion to enable female labour market participation. The • Christian People’s Party and the Centre Party, on the other side, stressed pedagogical and social aspects of day-care. Speakers from these parties only sporadically mentioned the need for labour supply. Female labour participation was not seen as relevant for all women, as it was thought that mothers worked outside the home only when required to due to individual economic necessity. Instead, the pedagogical advantage for children in collective day-care was underlined as was women’s desire to take care of their children at home (Innst.O.nr. 69 1974-75). Likewise, parties’ argumentation about the role of associational providers indicated coming conflicts. The Labour Party and Socialist Party saw associational provision as a supplement to public provision, as long as public day-care was not able to meet demand.67 The Christian People’s Party, on the other side, strongly advocated the particular role of associations. They mentioned associational provision most often and emphasised the usefulness of associational engagement in terms of “smooth, effective and human” operation.68

91 1980s and 1990s These diverging political party preferences, which underlied the compromise of the Day-Care Act of 1975, led to conflicts when day- care provision, especially religious-conservative associational day- care provision, developed not according to partisan expectations. First, day-care provision had developed mainly in the area of short- time day-care, which was especially the case for associational day- care provision (for operational parameters of different day-care providers see table). Second, job applicants for associational day-care complained that they had been questioned about their confessional views. Third, parents had started to complain about the religious content of day-care education provided to their children by some associations. On one side, the Labour Party worried that “many municipalities give away day-care extension, which is clearly a municipal responsibility, to private Christian institutions.” “This implies in many cases giving away expansion to organisations that, in principle, are against day-care and in the first instance are preoccupied with guiding principles and prioritising of part-time day-care […].”69 The Labour Party regarded specified demands by day-care operators concerning the personal convictions of their personnel as an “occupation ban” and not in accordance with international labour legislation.70 Instead, they wanted to ensure “free choice of occupation” Furthermore, the Labour Party felt that the passed formulation of guiding principles of the 1975 Day-Care Act made clear that “preaching and teaching are not in accordance with the intentions of the law.”71 In sum, the Labour Party feared that “a great and important welfare reform [was] going to waste.”72 On the other side, the Christian People’s Party defended the freedom of associations to provide day-care according to their own intentions. They should have the freedom to shape their provision, including free selection of personnel and free shaping of the curriculum. It should be self-evident that “personnel who take over the pedagogical responsibility for a facility […] are loyal and will conduct their functions according to the facility’s principles.” “Concerning private day-care providers, it is clearly manifested that they should have the freedom to constitute the course of their day-care facilities themselves. […] the Christian People’s Party is for its part fully prepared to take up a struggle to safeguard associations’ rights.”73

92 In 1980, the Labour government proposed amendments to the Day- Care Act of 1975, implying tightened statutory regulation over day- care (St.meld.nr. 93 1980-81). The Day-Care Act of 1975 had allowed associations to establish their own guiding principles in addition to general official guiding principles. The Labour government suggested introducing a differentiation between “general” and “non-general” day-care facilities whereby general day-care was associated with general educational content according to officially formulated guiding principles, and “non-general” day-care facilities had more freedom to set their own, additional, guiding principles. The Labour Party specified that the content of general day-care should correspond with “common beliefs” and a “tolerant attitude towards different worldviews.” Children’s questions about beliefs and norms should be “met with respect for different value orientations of parents and society” (Innst.S.nr. 367 1980-81). In addition, operators of general day-care facilities should not be allowed to ask job applicants about their personal worldviews. The crucial aspect of the proposed changes was that only “general” day-care facilities should be entitled to the whole amount of state • grants, while “non-general” day-care facilities should be restricted to 50% of state grants. Additionally, the Labour government introduced an overhaul of day-care subsidies by administrative law, which raised subventions for long-time day-care for younger children in relation to short-time day- care for older children. The ministry “[…] wanted to strengthen the expansion of half- and full-time day care […]” and “[…] day-care places for children at the age of 6 months to 1½ years […]” because “[…] parents wished […] that day-care would offer supervision of children while both parents work outside the home” (St.meld.nr. 93 1980-81).74 Right-wing parties rejected the intentions of the Social Democratic government (Innst.S.nr. 367 1980-81). They commented that “pedagogical and methodological freedom almost disappears. The so- called non-general day-care providers receive freedom that is conditioned on such a strong economic self-commitment that [it effectively limits the area to only a small number of providers].” Instead, they wanted “to stress that owners of private day-care facilities should have the freedom to recruit personnel that share the guiding principles established by the facility for its activities. The demand for qualifications may be the only aspect that can be audited

93 of others.”75 Thus, public authorities can only request of associations that associations should employ professional personnel, but not that associations disregard their preferences concerning the personal convictions of their employees. Shortly afterwards, when they gained power in 1981, right-wing parties clarified that “owners that have established a wide interpretation of guiding principles cannot be subjected to economic restrictions.”76 Instead, the right-wing alliance introduced two major amendments to the Day-Care Act of 1975 (Ot.prp.nr. 75 1981-82). One amendment would widen officially accepted guiding principles of associational day-care facilities. Originally, the Day- Care Act of 1975 obliged all day-care facilities “to develop their activity in close connection with children’s parents.” This aspect was abolished in favour of a right for associational self-regulation concerning educational content.77 The argument was that “there may be a need for a wider degree of freedom for some day-care providers, which are established with a more direct religious target. Many parishes of the State Church and private organisations and parishes provide day-care, which, in addition to a social orientation, are intended to teach clearly Christian content.” “If parents apply for places in day-care facilities with such statutes, the limitation on operational principles no longer exists, which is implicated by [the prescription] ‘in collaboration with children’s homes.’ ”78 In other words, associations were allowed to disregard parents’ wishes, if such would have implied a breaching with their preferred guiding principles. A second amendment made by the right-wing government widened associational influence on personnel recruitment. Associations, including associational day-care providers, were freed from the restriction in labour legislation that prevented employers from asking applicants about their personal beliefs and worldviews. Associational providers won the right to ask applicants about their convictions and make selections on the basis of their responses. “[…I]t is allowed to asked applicants about their convictions where it follows from the occupations’ character or where the facility has a specified guiding principle of, for example, Christian values and traditions.” Additionally, the right-wing government lowered subventions for long-time day-care for younger children, which was requested by the Christian People’s Party so “[…] that the largest possible number of

94 children is able to benefit from the pedagogical and social opportunity that day-care provides […]” (SF 1982 21.april).79 The amendments made by the right-wing alliances remained unchanged in the following decades, which can also be related to an ongoing right-wing majority in parliament (see Appendix A table A21 for government composition). However, when in power from 1986 to 1996, the Labour government was able to raise subsidy levels for long-time day-care and day-care for younger children by means of administrative laws, which do not need to be passed in parliament. Furthermore, the Labour government introduced extra subsidies to support a transformation of existing day-care facilities towards provision of longer opening hours and services for younger children from 1991 on (in connection to the lowering of school age from 7 to 6 during the first half of the 1990s, see Appendix A table A15 for subsidies). Subsequent right-wing governments abolished the extra subsidies for transforming day-care provision, and they overhauled the subsidy system, implying a return to the former relation of higher subsidy levels for short-time day-care and day-care for older children (in • connection with the introduction of home care benefits, see Appendix A table A15).

The “day-care settlement” of 2002 In 2005, a new Day Care Act was introduced by an alliance of the Labour Party, Socialist Left, Centre Party and Progress Party, who agreed upon a “day-care settlement” out of opposition in 2002. The right-wing government of the Christian People’s Party and the Conservatives agreed in principle upon the proposed changes (see section 3) but voted against the introduction of maximum levels on parent fees. However, the alliance of the “day-care settlement” enforced maximum levels for parent fees, going against the ruling right-wing government (Innst.O.nr. 128 2002-2003). This regulation was binding for all day-care providers, but was regarded as posing greater difficulties for associational providers. In sum, associational participation in day-care provision was accepted across party lines and statutory subsidies were allowed for unanimously (see table 7). Nevertheless, the discussions about guiding principles and “general” and “non-general” day-care show that associations’ freedom to regulate their day-care operations were not

95 agreed upon between parties. Furthermore, opinions about subsidy levels were varied among the different parties. On the one side, the Labour Party proposed lowered subsidies for day-care with “non-general” content and enhanced subsidies for long- time day-care and younger children. On the other side, the Christian People’s Party and their right-wing coalition partners defended or extended associational self-regulation of their day-care operations with full eligibility for public subsidies. Furthermore, they subsidised short-time day-care for older children more generously.

Table 7: Conflicts around associational day-care provision, Norway Policy items Supporting parties Rejecting parties Day-Care Act General guiding principles for all DNA, H, SP, SV KRF 1975 day-care providers in accordance with humanism, general ethics, democratic values, human rights, etc. Day-care provision (curricula, DNA, H, SP, SV KRF opening hours, etc.) in accordance with parents’ wishes Administrative Proportionally increased funding for DNA law, 1981 long-time day-care and younger children Amendments Rights for Associations: KRF, H, SP DNA, SV to the Day- 1.Widened rights to establish self- Care Act determined guiding principles 1975 2. Exemption from labour legislation on personnel selection when self- determined guiding principles applied 3. Full eligibility for state subsidies, irrespective of guiding principles Administrative Proportionally increased funding DNA laws, for long-time day-care and younger 1987-1997 children and extra subsidies Administrative Proportionally increased funding for KRF, H, SP,V laws, short-time day-care and older 1998-2005 children “Day-Care Maximum limits for parent fees DNA, SV, FRP, SP KRF, H, V settlement,” 2002 Day-Care Act, Municipal governance of All parties 2005 associational provision (opening hours, age of children, placing of children) and right of associations to municipal subsidies in addition to central state subsidies (80%) Labour Party (DNA), Christian People’s Party (KRF), Conservatives (H), Centre Party (SP), Socialist Left (SV), Liberals (V), Progress Party (FRP)

96 Summary In Norway and Sweden, political parties made controversial proposals and decisions about associations’ subventions and operational independence. In Sweden, social democratic governments limited subsidies for short-time day-care in 1963 and 1966, whereupon associational provision further decreased. In addition, social democratic governments precluded associations from obligatory municipal day- care planning and from short-time funding in 1973. Furthermore, they delimited the number of eligible associations during the 1980s and precluded them from extra subsidies for municipal employees. The obstacles for associational day-care, which were introduced by social democratic government, could eventually have postponed or slowed the growth of associational day-care seen since the middle of the 1980s in Sweden. In Sweden, right-wing parties protested against the aforementioned policies of social democratic governments. By contrast, they made all associational provider types eligible for statutory subsidies, equalised associations with public providers with respect to eligibility to • statutory subventions and, in addition, made associations eligible to municipal subsidies besides statutory subsidies during the 1990ies. If these more generous regulations had been implemented in the 1980s, these policies would arguably have led to earlier and faster growth of associations. In Norway, the Social Democratic Party intended to categorise associational providers as “general” and “non-general” in 1981, whereby “non-general” associations with specified curricula would be eligible only for a portion of statutory subsidies. Furthermore, they altered the subsidy system, which had favoured short-time day-care for older children. First, they reduced subventions for short-time day- care for younger children in relation to subventions for longer-time day-care and care for younger children in 1981 and again, when returning to power in the middle of the 1980s. Furthermore, they introduced extra subsidies to facilitate a transition towards longer opening hours and day-care for younger children. These changes to subsidy conditions can be regarded as at least supporting changes to the composition of associational provision since the middle of the 1980s, which implied fewer associations providing short-time day- care for older children, which were predominantly organised by

97 associations with specified curricula, such as churches, lay organisations and the Housewives’ Federation. The Christian Democratic Party and other right-wing parties, on the other side, opposed the above-mentioned social democratic decisions. Instead, they tended to make decisions supporting maintenance of or increases to associational provision levels. In 1981, a Norwegian right-wing alliance expanded the rights of associations to decide independently on their curricula and personnel. Furthermore, they maintained eligibility for statutory subsidies for “non-general” associations, which the previous social democratic government had intended to limit. Additionally, the right-wing alliance again lowered subsidy levels for long-time day-care for older children, a shift that was especially demanded by the Christian People’s Party. In sum, these right-wing decisions may have maintained the composition of associational providers until the end of the 1980s, which was dominated by religious-conservative associations providing short-time day-care for older children. In sum, social democratic governments made decisions about public funding and operational independence for day-care associations that reduced the eligibility and independence of (particular) associations. These decisions could be argued as having negatively influenced the overall level of associational provision in Sweden from 1963 to 1992, and as having contributed to altering the composition of associational providers in Norway since the middle of the 1980s. Christian Democratic and right-wing governments provided for (particular) associations more generously with public subsidies and allowed them more operational freedom, which would have positively influenced the level of associational provision in Sweden since 1992 and the level of religious-conservative provision in Norway from 1981 to 1990.

6. Parties and associations: Christian People’s Party, Housewives’ Federation and denominational associations As argued above, right-wing parties had been more supportive of statutory regulative settings that were favourable for day-care associations. Among the various right-wing parties, the Norwegian Christian People’s Party was very active in the Norwegian parliament

98 in supporting resource transfers and regulation competence towards certain types of associational providers. The Christian People’s Party influenced Norwegian day-care policy as the first party to defend the rights of associations to operate according to self-determined guiding principles, with religious content and personnel specifically recruited for this purpose. Furthermore, they defended full subvention for associations with self-determined guiding principles and supported generous subventions for short-time day-care. The Christian People’s Party is therefore seen as a central advocate of associational welfare in Norwegian day-care. According to the idea proposed here, political parties pursue favourable policies for those associations with which they share “congruent” constituencies. Thus, the Christian People’s Party is supposed to have pursued its policies due to its social congruency with the involved associations. The character of the Christian People’s Party policies, which defended generous subvention for short-time day-care and associational independence concerning recruitment of applicants with denominational convictions, is here interpreted as policies especially favourable for “religious-conservative” • associations. Therefore, the constituencies of associations belonging to this category, such as the Housewives’ Federation, Vesterlandske Indremisjonsforbund, Det Norske Misjonsselskap, Det norske Lutherske Indremisjonsselskap, Den norkse Santalmisjon and parishes of the Norwegian State Church,80 will be compared with the constituency of the Christian People’s Party for evidence of eventual congruency. The constituencies to be investigated further are the voters of the Christian People’s Party and the members and clientele of the associations; the clientele is indirectly measured by the prevalence of day-care facilities run by these associations.81 Due to a lack of individual-level data on the membership and clientele of religious-conservative associations, regional-level data on voting variations, membership and facility density are compared instead.82 The regional comparison will mainly be made between the Southwest of Norway and the rest of the country; the two areas are treated as contrasting cases.83 The idea of using these regions as contrasting cases is based on former studies, which showed regional differences between them in the degree of individual religiosity, election results for the Christian People’s Party and the density of a

99 conservative organisational “flora” (Øidne 1957, Rokkan and Valen 1964, Selle and Øymyr 1995). Øidne was first to highlight variances of religiosity between different regions in Norway. He emphasised the high degree of religiosity of the Southwest region of Norway in comparison with other regions of the country. Rokkan and Valen further scrutinised Øidne’s historical description by analysing the results of national elections, examining variations in voting patterns in the Southwest compared with other Norwegian regions. A main interest for their analysis was regional variations in voting patterns for the Christian People’s Party. Like Øidne, Rokkan and Valen concluded that the Southwest of Norway is distinct in that it is more supportive than other parts of Norway of the Christian-oriented Christian People’s Party. Studies of the “organisational flora” in various regions have likewise concluded that the share of religiously oriented or denominational organisations is higher in the Southwest than in other regions (Selle and Øymyr 1995). Organisations with either Christian- oriented activities or outright missionary organisations outnumber other forms of organisations, such as, for example, economic organisations, in the Southwest (see table 8).

Table 8: Types of organisations, share of all organisations by region

East-N Southwest-N North-N Buskerud Hordaland Finnmark 1941 1988 1941 1988 1941 1988 Christian orientation1 44 24 58 36 33 13 Missionary organisations2 27 10 44 15 16 5 Economic organisations2 25 13 13 12 27 15 (unions, craft, business organisations) 1analytic categories 2 main categories Source: Selle and Øymyr 1995

In sum, the Southwest can be said to have a specific culture of religious activism. Due to its specific religious character, the Southwest is here selected for the regional comparison as an area that can be contrasted with all other areas in Norway, because the protagonists of associational day-care in Norway seem to have

100 assisted day-care provision out of a common interest in religious- conservative stances on day-care provision. I will now turn to a regional comparison of member and voter densities of the Christian People’s Party and associations between the Southwest and other areas of Norway. In addition, some background information is provided concerning these organisations’ regional origins and normative stances. Afterwards, I will turn to a regional comparison of facilities and votes on both the county and municipal levels in order to estimate if clientele and voters of these organisations also have a higher density in similar geographical areas.

Regional distribution of constituencies and normative orientations Christian People’s Party The Christian People’s Party was founded by lay organisations in the Southwest of Norway in 1933.84 From the beginning, their party program declared that “the Christian People’s Party wants to preserve Christian and national values in church, school, working life and in • our culture,” which has since been announced as their guiding principle (Fjerstad 2001).85 Since the time of its foundation, the Christian People’s Party could rely on strong voter support from the Southwest region of Norway. Their share of votes is usually considerably higher in the Southwest than in other regions; they have more than 15% of the vote in the Southwest while less than 10% in other regions (see table 9, also Aardal and Valen 1997, Urwin 1997).

101 Table 9: Christian People’s Party (KrF), election results, shares in %1

County 1961 1973 1981 1989 2001 Östfold 3 3 2 2 3 Akershus 2 2 1 1 2 Oslo 2 2 1 1 2 Hedmark 1 1 1 1 2 Oppland 2 2 2 2 3 Buskerud 2 2 2 2 2 Vestfold 2 3 2 2 3 Telemark 3 3 3 3 3 Aust-Agder (SW) 4 4 4 4 5 Vest-Agder (SW) 3 5 4 4 5 Rogaland (SW) 4 5 4 4 4 Hordaland (SW) 4 4 3 3 4 Sogn og Fjordane (SW) 4 5 4 4 4 Möre og Romsdal (SW) 5 5 4 4 4 Sör-Tröndelag 1 3 2 2 2 Nord-Tröndelag 2 2 2 1 3 Nordland 3 3 2 2 3 Troms 2 3 2 2 3 Finnmark 1 2 2 1 3 National average share of 9.6 12.3 9.4 8.5 12.4 votes, % 15: 20% and over, 4: 19.9-15%, 3: 14.9-10%, 2: 9.9-5%, 1: 4.9% and under Source: NSD

Denominational organisations Missionary organisations were responsible for many denominational day-care associations in Norway.86 These organisations predominantly originated in lay movements from around the turn of the 20th century, which had their regional stronghold in the Southwest region (Øidne 1957, Rokkan and Valen 1964).87 A main task of these lay missions was to proselytise their compatriots to a more conscious Christian way of life.88 Deacon work was regarded as an important part of their mission, whereby they combined social and missionary aims. Normisjon, for example, has deacon work and care as one of its main four tasks, whereby their employees are requested to follow the protestant Lutheran conviction and obliged to ascribe to the values of the organisation (Det norske lutherske Indremisjonsselskap, today

102 Normisjon).89 Likewise, the Western Mission Union announces that “[our] employees have to a high degree combined preaching and social work” (Vestlandske Indremisjonsforbund).90 As can be seen in the table below, missionary organisations can rely on higher affiliation rates and a stronger disposition to do voluntary work among the population of the Southwest than in other parts of Norway. The Norwegian State Church is to a stronger degree a “national” organisation than the above-described missionary associations, because it has a member base of over 80% of the national population. However, only a portion of its members can be regarded as actively religious, and varied levels of activity can be observed between different regions in Norway. Again, the Southwest region is distinct from the rest of Norway in terms of higher levels of religious activity within the Norwegian State Church, including church attendance and voluntary work (see table 10). In sum, Christian denominational organisations can rely on higher activity rates of the population in the Southwest than in other parts of Norway.

• Table 10: Membership in missions and free churches and individual religious activity, 1990s Southwest Norway not All of Norway including the Southwest Home Mission membership1 4 1.8 2.3 Mission membership1 2 1 1.3 Free church membership 7.1 4.9 5.6 Voluntary organisation within 6.6 5.6 State Church Voluntary Christian work2 7.7 5.1 Church attendance (at least 16.5 4.2 8.1 monthly) 2 1 1990 2 not specified separately for the Norwegian State Church and other denominational organisations Source: NSD Den Norske Kirke statistikk (for 1990), other data: Botvar and Aagedal 2002

The Housewives’ Federation The Housewives’ Federation was founded in 1915 in Oslo, not in the Southwest like the other organisations investigated here. However, even though the Housewives’ Federation was a city phenomenon in its early beginnings, the organisation’s predominance in the cities vanished by the 1920s and 30s, when it spread throughout the country.

103 It was mainly in the Southwest that they gained new members, and membership density in the Southwest soon exceeded membership density in the original centre around Oslo (see figure 7 and Appendix A table A23).

0,9 e, %-%

0,6

0,3

0 1930 1950 1970 1980 1990 2000

-0,3

-0,6 differencemembership regional of density from averag

Oslo area Southw est Norw ay (w ithout Oslo and Southw est)

Figure 7: Housewives’ Federation’s membership densities by region Source: see Appendix A table A23

The main normative orientations of the Housewives’ Federation could be described as a combination of modern and conservative values. On one side, they aimed to professionalise the role of housewives and mothers through fostering scientific approaches to raising children and rationalising housekeeping. Examples of their activities are the provision of informational material and the organisation of courses and fairs on the topics of nutrition, household machines, insurance, etc. (Solbraa-Bay 1970, 1985). Furthermore, they advocated day-care along the ideas of modern pedagogy, whereby professionally led part- time day-care was seen as the best solution for children (NOU 1972:39, Nygårdsmoen 1998, Lien 1989, Housewives’ Federation’s annual reports; comment on NOU 1984:20).91

104 On the other side, they advocated conservative roles for women, as indicated in their organisation name, “Housewives” (which they abandoned in 1998). They also committed themselves to follow Christian values, which is one of their main guiding principles: “§ 2 The Housewives’ Federation will work along Christian principles” (Solbraa-Bay 1985, Housewives’ Federation’s annual report 1997).92 In sum, all organisations involved in provision of religious- conservative day-care have their social strongholds in the Southwest region of Norway (see figure 8). Furthermore, they shared a commitment to Christian religiosity as an ideological common denominator and announced their Christian-religious orientation among the first paragraphs of their guiding principles. Another shared normative orientation of the Christian People’s Party, the Housewives’ Federation and denominational providers is their positioning in recent conflicts about family life and child education. One conflict centred around home care allowances, which are given to parents who stay at home with their children, introduced in 1997. Another conflict occurred around the introduction of a prescriptive Christian orientation for all (public) day-care in 1983.93 The Christian • People’s Party was, in both cases, the political initiator for these measures, and was supported by the Housewives’ Federation and various religious groups (Housewives’ Federation’s Annual report1995, comment to government investigation, Annual report1995, National congress 199394). Table 11 summarises aims and regional-social background of the organisations investigated here.

Table 11: Aims and regional backgrounds

Regional Regional Basic programs Proponents for origin stronghold A Home care allowance; B Christian education for public day-care Housewives’ Oslo; Oslo; Housewives’ A Federation Southwest Southwest Federation B interests, Christian values Denominational Southwest Southwest Christian B organisations denomination, Missionary Christian Southwest Southwest Christian values A People’s Party in society B

105 Dispersion of associational day-care provision Regional level As the table below shows, denominational and conservative associations operate day-care facilities more often in the Southwest than in the rest of the country. In 1980, the share of these associations’ facilities was nearly twice as high as in the rest of the country. Denominational and conservative organisations operated 45% of all day-care facilities in the Southwest, but 24% in the rest of Norway in 1980. The same pattern of associational day-care provision remained apparent in 1990 and 2000. Then, denominational and conservative organisations operated 36% (10%) of all day-care facilities in the Southwest, but 25% (7%) in the rest of the country (see table 12).

106 •

Table 12: Norway: day-care facilities by region and ownership, shares of all facilities operated, % Public Associations Religious- Others N organisations conservative (sum) associations (sum) M PO1 PD1 E P I M1 1980 Southwest 40 33 12 45 3 6 Na 6 15 798 Rest of Norway 60 17 7 24 3 10 Na 5 18 1755 1988 Southwest 51 27 10 37 1 12 Na Na 13 1322 Rest of Norway 66 14 6 20 2 13 Na Na 15 2721

107 1990 Southwest 51 28 8 36 1 3 4 6 14 1492 Rest of Norway 63 20 5 25 1 4 3 5 13 3156 1996 Southwest 48 5 8 13 2 13 15 9 39 1973 Rest of Norway 48 2 5 7 3 17 17 8 45 4434 2000 Southwest 53 3 8 10 2 12 14 9 37 1811 Rest of Norway 51 3 4 7 3 15 16 8 40 4019 1 categories changed over time (see Appendix A table A20) M (municipalities, counties, state), PO (private organisations), PD (parishes, denominational organisations), E (employers), P (parents), I (individuals), M (miscellaneous) Source: NSD, own calculations

Not all forms of associational day-care had higher shares in the Southwest counties, but only these specific ones. Other non-public day-care operators had either lower or similar shares of facilities in the Southwest. Parent cooperatives, for example, had lower shares of day- care facilities in the Southwest than in the rest of Norway. In 1980, they operated 10% in the rest of the country, but 6% in the Southwest. In 1990 and 2000, parent cooperatives provided still more day-care in other parts of Norway than in the Southwest. The interpretation here is that not a general interest in non-public day-care provision per se, but rather particular interests in religious- conservative day-care explain the higher shares of associational day- care of denominational associations and private associations in the Southwest. Figure 8 summarises the regional variations of partisan and associational constituencies in 1980, which all tend to be higher in the Southwest than in the rest of the country.

'religious- 50% conservative' f acilities; 45%

40% 'religious- conservative' 30% f acilities; 24%

20% KRF votes; 16%

Mission 10% Housew ives 6% membership; 3,6% membership; 1,34% 1,8% 0,62% 0%

Norw ay w ithout Southw est Southw est

Figure 8: Religious conservative facilities, KRF (Christian People’s Party) votes, Missions’ and Housewives’ Federation’s memberships, shares by region, %, 1980 Source: NSD, NOS Population and Housing Census, Housewives’ Federation

108 Municipal level Southwest Norway and the rest of Norway are very large areas of comparison. In order to make a more reliable estimate of congruencies between associations’ day-care provision and Christian People’s’ Party, information at a lower aggregation level will be looked at. The lowest possible aggregation level of statistics concerning day-care provision is the municipal level, which is now chosen as the level of analysis. A comparison is made between the relative level of religious-conservative day-care provision in the municipalities and voter support for the Christian People’s Party in the municipalities (information on associations’ membership shares is not available at the municipal level). Table 13 shows the relative number of municipalities that have certain shares of religious-conservative day-care provision and certain shares of the Christian People’s Party vote.95

Table 13: Religious-conservative associations and Christian People’s Party, 1980 Municipalities with… • Religious-conservative day-care facilities of all facilities, shares, % 2 None Low High % of Municipalities municipalities (N) Christian Lowest 63 22 15 100 104 People’s Low 52 36 12 100 109 Party High 53 19 29 100 108 Votes1 Highest 37 14 50 100 103 Municipalities 217 96 111 100 424 (N) Missing 30

1 Christian People’s Party’s vote share in a municipality: highest = 20.5%-, high =11.2%-, low = 3.45%-, lowest = - 3.45% 2 religious-conservative day-care facilities as a share of all facilities in a municipality: high = 50% -, low =-50%, none = 0 Source: NSD

The overall impression is that religious-conservative associations provide larger shares of day-care in those municipalities where the Christian People’s Party enjoys higher voter support. In contrast, other associations do not provide more day-care in municipalities where the Christian People’s Party have high voter support (see Appendix A table A24).

109

Summary It seems that associations and the Christian People’s Party share rather congruent constituencies. This is indicated by shared cultural values, higher voter support for the Christian People’s Party and higher membership densities and higher provision rates of specific day-care associations in the Southwest.

7. Day-care: Summary of main findings The main idea of this thesis is that statutory support for welfare associations is moulded by political parties in accordance to their particularistic interest in associations with either more or less “congruent constituencies.” The examples in this chapter have been Norwegian and Swedish parties’ politics concerning associational involvement in day-care provision.

I. Partisan day-care policies The findings were, first, that day-care policies of parties diverged concerning the relative role and subvention of associations, which could arguably be regarded as having shaped associational provision. Social democratic parties tended to restrict associational day-care provision either by diminishing subsidies for associations or by enhancing public control over the operation of day-care provision. Social democratic parties reduced associations access to subventions in Sweden during the 1970s and 80s, and reduced subventions for associations by lowering subsidies for short-time day-care for younger children in Sweden during the 1960s and in Norway during the 1980s. In addition, the Norwegian Social Democratic Party introduced extra subsidies to encourage provision of day-care for younger children and longer opening hours during the 1990s. Furthermore, the Norwegian Social Democratic Party tried to enhance public control over associational providers in Norway (in Sweden, public control over associational provision was not a point of contention between parties). It may be argued that the social democratic influence on public day- care policies diminished associational provision in Sweden during the 1960s, 1970s and 80s and altered the composition of involved associations in Norway during the late 1980s and early 1990s.

110 By contrast, right-wing parties supported associational provision of day-care. In Sweden, they exerted pressure to make associations eligible for subsidies in the 1970s and the 1980s and introduced full eligibility of associations to all forms of statutory subventions in the 1990s, which was followed by a further increase in associational provision of day-care. In Norway, right-wing alliances objected to social democrats’ attempts to place limits on associations’ sovereignty and eligibility, and, instead, decreed in the 1980s that associations should gain full eligibility and broader rights concerning day-care operations, curricula and selection of personnel. In addition, they restored subventions for short-time day-care and older children in the 1980s and abolished extra subventions to transform day-care services in the 1990s. The growing strength of religious-conservative associations until the 1990s is interpreted as having been assisted by this policy. In sum, party politics differed concerning associational involvement. Social democratic parties were trying to limit the independence of associational operations and to reduce subventions for (particular) associations, while right-wing parties supported • independent operation and generous subventions for (particular) associations in Sweden and Norway. However, the relevance of party political influence on the level and composition of associational provision is limited.

II Parties’ and associations’ constituencies Variations in party politics are argued to have been partly in accordance with the notion that they are based on shared interests in congruent constituencies with the main involved associations. In Norway, the Christian People’s Party shared several social commonalities with a predominating set of associations, which were organised by missions, parishes and the Housewives’ Federation. The Christian People’s Party and these associations had their social origins and main social bases in the Southwest, as indicated by vote shares, membership density and day-care provision density. Furthermore, they shared a commitment to Christian values, which they also shared with the population of the Southwest region. The social similarities between the political party and religious-conservative associations are interpreted as incentives for the Christian People’s Party to use its political influence in favour of day-care associations. In contrast, the Norwegian Social Democratic Party’s stance on day-care associations

111 is seen as based on an absence of social links with large associational day-care providers. However, political parties’ diverging efforts to support associational day-care no longer correspond with associations’ social backgrounds in Norway and Sweden since the 1990ies.96 Although Swedish right-wing parties introduced full eligibility for associations, their efforts to establish associational provision were not met by an upsurge of associations that had close relations to these parties. Right-wing parties had argued for full eligibility for Montessori, Waldorf and Swedish State Church associations, but the share of services provided by these associations remained minor in comparison with parent cooperatives and employers. In Norway, the Christian People’s Party could rely on the efforts of missions, parishes and Housewives during the first decades after statutory day-care subvention. However, since the 1990s, the Christian People’s Party had only limited influence concerning associational provision as the number of religious-conservative associations decreased. Parent cooperatives, which are today a larger provider type than religious-conservative associations, can no longer be regarded as sharing the special day-care interest of the Christian People’s Party, as indicated by their provision of day-care for younger children and longer opening hours. Furthermore, parent cooperatives share no common social background with the Christian People’s Party in the Southwest, where they tend to be under-represented. In both countries, parent cooperatives became the largest associational type. Besides their organisational form, they seem to have no particular characteristics. Similar to municipal day-care, they provide day-care with long opening hours for children of all ages. Furthermore, they have no particular aims or normative orientation connecting them to a particular political party. They differ from other organisational forms in that they are small-scale organisations that allow parents a high degree of control over their children’s day-care. Associational day-care was not eagerly pursued by social democratic parties in Norway and Sweden. They did not push regulations to favour associational provision, but reluctantly accepted associational solutions. However, this lack of interest in associational provision is not a general attitude: In other welfare areas, social democratic parties were indeed interested in establishing associational provision. As the Christian People’s Party showed an enhanced interest in associational provision through associations with

112 coinciding social backgrounds and interests, social democratic parties also supported associations with closely related social backgrounds. An example of social democratic interest in associational provision is their policy towards housing cooperatives.

•

113 III. Housing in Sweden and Norway

1. Introduction: Housing The share of associational housing in Sweden and Norway ranks high in international comparisons (see table 2). A further difference to other countries is that the active associations are cooperatives in Sweden and Norway, while in other countries, such as the Netherlands, France, and Germany, employers, unions, philanthropies or churches provide rented housing (Power 1993, Union HLM, Haakfort et al. 2002, Gesamtverband der gemeinnützigen Wohnungsunternehmen). An argument for strong public subvention and regulation of associational housing growth in Sweden and Norway will be theme of sections 2 and 3. It will be argued that statutory intervention was an important factor for associational housing growth and the shaping of associational tenure by setting preferable preconditions. However, one has to keep in mind that states’ influence over housing provision is more restricted by market conditions and individual purchasing power than is the case for other welfare areas, such as education or income maintenance policies (Boelhouwer and Heijden 1992, Lundqvist 1992c; for elaboration of policy influences on tenure patterns, see Appendix C). In section 4, the main involved associations, i.e. cooperatives, will be presented in greater detail. In addition, statutory regulations concerning the tenure of cooperatives will be presented. Afterwards, it will be argued that political parties struggled to influence the formulation of public policies concerning associational involvement and that they pursued their policies out of particular partisan interests in the main involved associations. The interests of political parties will be described in sections 5 (concerning public policies) and 6 (concerning main associational providers). To provide a notion of general interests in associational housing provision, I will start by introducing functions and related historical examples of associational housing engagement.

114 Functions and associational interests in the area of housing The leading question for this introduction asks which interests associations have in housing provision. For this purpose, I will first describe functional aspects of tenure and ownership of housing. Which functions do the owners of the house or dwelling influence? What are the implications of this for the occupant or tenant97 of the dwelling? Second, I will give examples of associations dealing with ownership of housing and make some assumptions about their interests connected to the specific functions accompanying ownership.

Functions of tenure and ownership Housing has many functions for the occupant or tenant. A basic function could be called habitation or a “place to live.” In addition, various sub-functions are connected to housing – such as health, security, privacy, social contacts and property rights – depending on the quality of a dwelling, the environment of the dwelling and the ownership or tenure status of the occupant. Here I will focus on • functional implications that accompany ownership and tenure status of the occupant (see figure 9). Ownership and occupation usually come as a “bundle of legal rights” specifying the rights and duties of persons towards each other as owners or occupants of a dwelling (Marcuse 1994). The difference between owners and occupants, which do not have to be the same persons, is related to who provided capital for the property and who occupies it. In most countries, the legal specification of ownership status provides the owner with more control rights than the occupant concerning the usage and disposition of a dwelling (Lundqvist 1992c). First, owners can influence the usage of the property. They can, for example, decide about the length of the occupancy, about reconstruction, renovations or subcontractors for services and goods, etc. Owners can also impinge on the privacy of occupants, for example, placing restrictions on living habits such as smoking, restricting or forbidding pet keeping, limiting the number of persons living in the dwelling, etc. Concerning occupants other than themselves, control opportunities include rent setting, the selection and eviction of tenants, detailed prescriptions about the usage of the dwelling and eventual inspections of the dwelling.

115 Second, owners have the property rights to the dwelling, which entitle them to dispose of the dwelling, either through selling or bequeathing it. Besides providing wider opportunities for owners to make market gains, dispositions may affect occupants’ housing security due to changed owners and subsequent changes in conditions. Furthermore, if the (prospective) owner also acts as the investor, owners can, more directly than occupiers, influence the supply of dwellings. They can relatively directly influence whether and when investments in housing are made. Before the construction process starts, they can also influence the types of dwellings they are going to acquire, in terms of standards, space, partitioning of rooms, etc. In sum, ownership of housing provides owners with several control rights, including influence on occupants’ housing situation and the housing functions that come with occupation, such as privacy, security, housing standards, etc. The rights of occupants, on the other side, can be more limited than those of owners. Rights of occupants are usually called tenures, of which typical examples are tenancy in rented housing owned by landlords, owner-occupation in detached housing, owner-occupation in multi-dwelling houses (condominium ownership) and occupation in cooperatively owned houses. Owner-occupiers have owner-(like) rights towards a dwelling, which usually provides them with a high degree of control over their housing situation. Rental tenants on the other hand, have no owner rights, which can limit them in several ways. For example, they only indirectly influence the demand for dwellings, which can make it difficult for them to find suitable housing. They may have to accept less-than-ideal dwellings in terms of health, due to poor standards, privacy among tenants, due to inadequate space, etc. They also have to accept time- limited contracts or are threatened by eviction, reducing their housing security. In addition, they have to accept limitations on their privacy by owners, in the form of, for example, visits by the owner, limitations on the number of tenants allowed to share the dwelling, or prescriptions about living habits. Furthermore, renting tenants usually cannot influence administrative decisions, for example, about reconstruction or services, and they have limited influence over their housing costs.

116 In other words, renting-tenants have only indirect influence over their housing situation and are subjected to some influence by the owner of the dwelling. Between having complete owner rights and no owner rights, there are other forms of tenure that come with varying “degrees” of ownership. Owner-occupiers in condominiums, for example, share collective ownership of parts of a house, such as the roof or the site, which they have to manage in cooperation. They also have to make collective decisions about such administrative matters as supplier contracts and reconstruction decisions. Apart from such collective influence, which may limit the functionality of individual occupants’ housing situation, their tenancy rights come very close to those of single ownership. Cooperative occupation and ownership is entirely collective, i.e., single occupants do not own their dwellings alone, but occupants together own all dwellings. Privacy and security are greater for cooperative tenants than for renting tenants, due to rights to influence collective decisions. Nevertheless, cooperative owners are dependent on majority decisions and the financial soundness of the co-owners. • Their security can be threatened through collective insolvency, which can lead to a shift of ownership. In addition, their privacy can be limited through rules enforced by the majority. Certain cooperatives have, besides collective ownership of dwellings, collective affiliations in superordinate cooperative organisations. Superordinate cooperative organisations can impose regulations on local affiliated cooperatives, which can further limit the occupant’s private decisions, even if he can influence the decisions taken at the superordinate level. This was the case in Swedish and Norwegian superordinate cooperatives, which, for example, delimited through statutes individual occupants’ decisions about prices and disposal of their dwellings.

117 OWNER-OCCUPATION RENT-TENURE PRIVATE OWNERSHIP Home-owners Private landlords (Churches) (persons) Condominium owners

Co-op c Co-op a (farmers,middle classes, unions) Private landlords (companies, firms)

Co-op b Associational landlords (tenant-org,. unions, middle classes) (Employers, unions, etc.)

118

Public landlords

COLLECTIVE OWNERSHIP OWNER-OCCUPATION RENT TENURE

Figure 9: Ownership and tenures Co-op a: independent cooperative (collective ownership, occupant’s market free access and sale rights) Co-op b: affiliated cooperative (collective ownership, occupant’s access and sale rights delimited, membership in a superordinate cooperative) Co-op c: rental cooperative (cooperative management, but no ownership)

In sum, co-owner tenures come with some limitations for the occupants concerning decisions about the development, usage and disposition of the building. In comparison with owner-occupiers, who can directly make individual decisions, co-owner occupants’ influence is limited due to the influence of other owners over collective decisions. Owner-occupation of housing provides occupants with many aspects of control over their habitation in terms of security, privacy and functionality, while delimited forms of ownership come with limited control for the occupants. The advantages that go with popularity of home-ownership contribute to its popularity. However, home-ownership is generally limited to those who have access to the capital needed to purchase a dwelling. In contrast, rental tenancy has the advantage of a pay-as-you-go scheme, making it more affordable than home-ownership for those in the lower strata of society. Forms of co-ownership of houses fall between home-ownership and rental tenancy; they reduce the capital investment required of individual owners due to shared ownership of the property.

• Historical development of housing interests and ownership During the first half of the 20th century, urban housing was predominantly provided by private landlords, which owned and let the largest shares of urban housing.98 Dwellings were often of low standards, with poor hygiene, heating, lighting, etc. Furthermore, housing shortages led to overcrowding and sub-letting of rooms, which meant that households often shared their apartments with other persons. Around 15% of all apartments had more than 2 persons per room in Europe (ILO 1930; see Appendix A tables A26 and A27 for Norway and Sweden). In addition, evictions of tenants due to overdue rent payments were common (cp. ILO 1930, Strömberg and Pooley 1992). The effects of these housing conditions on health, family life, life conduct and security were increasingly regarded as a social problem affecting many groups in Europe (ILO 1930). Private landlords and private investors, who were the predominant investor and owner groups of urban housing, were seen as partly responsible for current housing conditions. Labour movement organisations especially saw free-market development and ownership as the main reason for housing shortages, poor conditions and high costs. However, other social groups also regarded private landlords

119 and housing investors as too narrowly pursuing profit interests and therefore often reluctant to meet the accelerating needs for housing in newly developing areas or reluctant to ensure “decent” housing standards and improve working classes’ life conduct (ILO 1930, Harloe 1995, Strömberg and Pooley 1992, Power 1993, Umrath 1953). In response to prevalent housing conditions and shortages, various associations (and public authorities99), for example, labour movement organisations, employers, philanthropic organisations and churches, began to intervene in the housing market (ILO 1930, Harloe 1995, Strömberg and Pooley 1992).100 In contrast to private landlords, associations usually had no direct for-profit interest in housing, but often chose limited-profit types of organisation to provide dwellings “at cost,” meaning at prices oriented to production costs. Their hope was that at-cost housing would lower the financial demands on occupants in comparison to the demands posed by profit-oriented private landlords and investors. Associations’ housing activities were thus closely oriented to interests in specific client categories. Employers, for example, were predominantly interested in solving the problem of insufficient housing for their labour force. They developed and provided housing in order to attract workers to newly industrialising areas close to their plants, where private investors were reluctant to invest. Examples of such employer-provided housing are in the German Ruhr region, Northern France, Dutch Limburg and small-scale industrialised areas in Sweden (brukssamhälle) (Blumenroth 1975, Strömberg and Pooley 1992, SOS Population and housing census 1933 and 1945). Employers eventually also worried about the living conditions of the labour force, concerned that unstable housing situations could lead to an unfit workforce and social unrest. Therefore, some employers engaged in outright social control through inspections and prescriptions for occupants living in housing they provided. For example, German plant owners applied “housing models” advocated by the German Society for Social Policy, of which they “expected worker families to learn how to ‘live correctly’ under the supervision of bourgeois social reformers; i.e., they were to adopt the virtues of order, self-discipline, and household economy” (Häußermann 1994). In the Netherlands, programs were adopted with an “emphasis on a management style which stressed the need to inspect, discipline, re-educate and control the potentially feckless tenants, who were to live in this housing” (Harloe 1995). In France,

120 the Society of Parisian Workers’ Housing Areas invented management and inspection via a “concièrge” (Power 1993). Other associations, such as philanthropic organisations, cooperatives, tenant organisations and churches, had a more pronounced “social” interest in providing housing than employers; that is, they wanted to improve housing for people in the lower social strata, who had so far lived in unhealthy or insecure housing conditions in densely urbanised areas. However, their “social interest” was connected to different expectations concerning their clients. Philanthropic landlords, for example, chose “morally decent” tenants and made control visits and detailed prescriptions about how tenants were expected to conduct themselves; should they not comply they could be evicted. Such forms of social control were used, for example, by philanthropic landlords such as Octavia Hill in London, who advocated “strict enforcement of rent payments and tenancy conditions.” The adopted model was to have rent collected by middle- class women, who also inspected the dwellings (Power 1993, Holmans 1987). Scandinavian cooperatives, for example, took a more liberal stance, • granting unlimited tenancy rights and offering courses, expeditions and social evenings for their tenants to foster a community spirit (Annaniassen 1996a, Bengtsson 1993, Lundevall 1994). They also used their organisational position to contract suppliers that were socio- politically sympathetic with their organisation, such as construction worker cooperatives or trade union insurance firms (Lundevall 1994, Annaniassen 1996a). In sum, associations commonly engaged in housing provision to overcome housing problems and often provided housing at cost, but with varying emphases according to their specific interests. They provided housing for specific client groups and attempted to influence the “well-being” of occupants via contact and control. Furthermore, they eventually used their ownership position to deliberately choose specific subcontractors for their housing investment. However, associational housing was usually limited (see Appendix A table A4). The next section will examine the relation between associational efforts and public housing aid in order to discuss the capability of associations to provide housing without public support in Sweden and Norway.

121 2. Associational limitations and statutory intervention This section examines the extent to which associations have been capable of providing housing without statutory assistance in Sweden and Norway. The basic assumption is that associations are unable to provide housing at a substantial level through their own efforts, but depend on public subvention. To estimate associations’ capabilities, associational provision is traced back to a time before major public intervention occurred. However, due to limited information about ownership in housing statistics before state intervention, “pure” self- sustained associational provision is difficult to grasp. Data used to measure associational provision are, first, census data on the ownership structure of housing stocks and, second, statistics about new construction activity by investor categories. To assess statutory support for housing associations, the overall level of loan subvention is described, as is the distribution of support to various investor categories, looking at the coverage of investor categories’ housing projects by statutory loans and the coverage of their project costs. The earliest national housing census in Norway dates from 1920. It indicates an absence of associational housing and more than 90% private ownership, but should be interpreted carefully (Gulbrandsen 1988). Subsequent housing censuses give no specific information about ownership before 1960.101 Swedish statistics provide relatively secure information since 1945, when the first nationwide and comprehensive housing census was conducted. It states that housing associations had a share of about 13% of all dwellings: private employers owned 7% of all dwellings, foundations, 2%, and cooperatives, 4% (83% were privately owned, 4% were publicly owned; SOS Population and Housing Census 1945). Based on an earlier census, which excluded parts of the rural area and remained unclear in categorising associative housing, it could be assessed that associations owned a share of around 8% of the housing stock in 1933, of which employers owned about 5% and philanthropic organisations, shareholder companies and cooperatives without profit orientation about 3% of all dwellings. Furthermore, data indicate that associational housing increased from about 8% in 1933 to 13% in 1945, of which employer housing increased from 5% to 7%, while

122 other forms of associational housing increased from about 3% to 6% (cooperatives, foundations, etc.).102 To what extent did associations accomplish these levels (and eventual increases) of housing provision by their own efforts, that is, without statutory intervention? In order to estimate the influence of statutory subvention on associational housing, the levels of statutory intervention and associational construction activity are compared. Two phases of high state intervention can be distinguished (see Appendix A table A28). A first high intervention phase occurred from 1917 to 1923, when statutory aid went to 45% of all newly built dwellings.103 A second high intervention phase lasted from 1941 to 1945, when 70% of all new dwellings were built with statutory help. In the years of low intervention, 1924 to 1940, statutory aid assisted about 15% of all new dwellings (ILO 1930, Silk 1948, SOS Population and Housing Census 1945). The overall level of public intervention can be compared with the development of associational construction activity. Public utility (‘allmännyttiga’) companies, owned by public authorities and/or associations, built 40% of all new dwellings during the first high • intervention phase from 1917 to 1923. In contrast, from 1924 to 1940, when statutory intervention was low, public utility companies built on average 13% of all new dwellings. During the following high intervention period, from 1941-45, public utility organisations built 33% of new dwellings. Associational construction activity can only be estimated due to the semi-public character of public utility companies, but was about 21%, 12% and 27% in respective periods. It remains unclear whether public subvention was targeted to public utility companies, because legal regulations did not inhibit subvention for private investors.104 However, subvention policies were often formulated expressing preferences for public utility investors (e.g., loans in the 1920s, SFS 1922:413, SFS 1929: 127). More detailed information about the de facto distribution of public aid105 to different investor categories is available for the period from 1931-45. Public subsidies went to 33% of all newly built dwellings for the entire period from 1931 to 1945 (see table 14). Public subsidies and loans were not distributed evenly among the various investors, but were predominantly granted to associations, specifically cooperatives and public utility companies, whereby public utility companies106 received loans for around 88% of their dwellings and cooperatives

123 received loans for 69%. Private investors, on the other hand, were granted loans for only 20% of their construction projects. Thus, cooperatives and public utility investors received a larger portion of public support.

Table 14: Share of publicly supported new dwellings by owner type, %, 1931-45, towns with more than 1000 inhabitants 1931-35 1936-40 1941-45 All years Private person 8 6 57 19 Firm, real estate company, 9 6 44 22 employer Cooperatives 50 31 89 70 Public utility 84 86 91 88 State, county, municipality 4 8 40 23 All 16 13 69 33 Source: SOS Population and Housing Census 1945

That public aid was important for associational construction activity can be derived from the pattern of increases in aid and construction activity during this period. As is shown in table15, increases in public aid were followed by increases in associational construction activity: The level of public support for cooperatives and public utility investors increased from around 50% from 1931 to 1940 to 90% from 1941 to 1945, whereby the share of associational construction activity simultaneously increased from 9% to 26%.

Table 15: Construction activity, 1931-1945, by investor, % State, county, Public utility Cooperative Employer Private Private municipality companies, landlord owner foundation 1931- 3 3 6 1 66 18 1940 1941- 6 3 23 1 53 14 1945 Source: Byggnadsverksamheten 1933, 37, 40, 45

A further indicator of public subsidies’ influence on associational (and public) housing achievements is the age of dwellings owned by them. Sixty-nine per cent of the dwellings owned by cooperatives and public utility companies in 1945 were new, built between 1931 and 1945.107 This was a period when they received 70-88% of public aid. From 1941 to 1945, when associations received statutory aid for 90% of their dwellings, they acquired 39% of their entire 1945 housing stock (see table 16). To sum up the developments of statutory intervention and associations’ construction activity, it could be said that levels of

124 public support and shares of associational construction activity co- vary over time. This indicates that public support has been a considerable factor influencing associations’ activity, although there may be some reservations concerning employers, which seem to have accomplished their housing provision with little public support.108 Employers already had 5% of the housing stock in 1933, when public intervention had been short-lived. Furthermore, they seemed to have increased this share to 7% in 1945 without receiving much in the way of public subsidies.

Table 16: Age of housing stock by owner, %, towns above 1000 inhabitants 1880-1930 1931-35 1936-40 1941-45 Housing stock 1945 Sum Total national Private person 69 10 14 7 100 69

Firm, real estate 69 7 13 11 100 16 company, employer Cooperatives 32 11 16 42 100 8

Public utility 28 14 27 31 100 3

State, county, 88 4 3 6 100 4 • municipality

All owners 65 9 14 12 100 100

Source: SOS Population and Housing Census 1945

Cooperatives, foundations and other associations, on the other hand, could increase their activity substantially when public support was provided. Construction activity among associations increased parallel to increases of public support for their construction projects. In addition, they acquired 30-40% of their total 1945 housing stock during the 1940s, when they received subsidies for more than 90% of their projects. Thus, it is assumed that associations’ housing stock share would have been considerably lower in 1945 if they had not received public support. The question of this section was whether associations had been able to provide housing on a substantial level without public help. Based on the data discussed above, it is argued that associations accomplished much of their provision through public help. The central role of public support is somewhat surprising in that associations might be expected to be more able to provide housing for social groups with regular incomes. In addition, associations could

125 been expected to compete successfully against private landlords in attracting such classes since they were providing housing at cost, which could have offered better housing standards for the same price as they would pay in the for-profit housing market. Inexperience in housing provision may have been one reason, but, as the growth of associational housing with statutory aid shows, associations seem to have been capable of developing the necessary organisational skills in financing, planning and management. The main difficulty seems to have been raising capital from private banks. Mortgage banks may have been reluctant to lend money to organisations that had no assets and had so far not proven capable of providing housing successfully on the market. Employers, on the other hand, who irrespective of public support provided housing with some success, would have an advantage in that they could rely on assets and/or be more trusted by the banks. So far, statutory intervention has been discussed with respect to associations’ ability to provide housing through their own efforts. In the following, statutory intervention after the Second World War in Sweden and Norway will be looked at with an emphasis on how housing policies were shaped concerning different investor categories.

3. Statutory intervention The composition of ownership and tenure of the housing stock underwent a profound shift beginning in the 1940s in Norway and Sweden. Private ownership of multi-dwelling housing declined, while the availability of at-cost housing increased. In Sweden, ownership of public and cooperative housing increased, and, in Norway, cooperative ownership increased (see tables 17 and 18, see Appendix A tables A29 and A30 for tenure and ownership of all houses in Sweden). Here, the main interest is the increase in cooperative housing, which is the main associational type of housing provision in Sweden and Norway. The question is whether statutory housing policies can be interpreted as having supported the tenure and ownership change and, especially, the increase in cooperative housing provision. To answer this question, the adopted housing policies will be described and discussed with a view to identifying the extent to which they favoured or disfavoured various investor and tenure categories. In order to

126 provide an overview of different policy measures, I begin with a brief discussion of a wider range of housing policies, such as land policy, taxation, tenure regulations, etc. Afterwards, the focus will be on capital subventions.

Table 17: Sweden: Ownership of multi-dwelling houses 1945-1990, shares, %

1945 1960 1970 1980 1990 State, municipality, 8 22 34 38 40 public utility, etc. Cooperative 10 21 24 27 28 Private person, 91 57 42 33 59 companies 2 2 Total 100* 100 100 100 100 * does not add up to 100 due to double counting (SOS Population and Housing Census 1960) Source: SOS Population and Housing Censuses

Table 18: Norway: Ownership and tenure of housing stock 1945-2000, shares, %

1950 1960 1970 1980 1990 2001 Cooperative tenure1 6 11 13 19 19 14 Rented 41 36 34 26 21 24 of which municipal 1 2 4 Owner-occupied 53 53 53 55 64 62 • of which 3 5 4 condominiums2

Total 100 100 100 100 100 100

1 “cooperatively owned” includes ownership in companies by shareholders, which are under roughly the same regulations as independent cooperatives 2 Estimations: the share of condominiums is included in the category “owner-occupied.” According to Gulbrandsen, condominiums were unusual before the 1970s. Source: NOS Population and Housing Censuses; NSD; Gulbrandsen 1988; Lundqvist 1992a

Occasionally, official policy aims for housing expressed favour or disfavour of tenure and investor categories. From the 1940s to the 1990s, according to officially stated policy aims, “speculative” construction and “commercial” investments were to be prevented, which can be interpreted as an expression of disfavour towards private landlords and private renting in Norway and Sweden (Innst.Ot.I. 1946; St. meld.nr.34 1988-89; SOU 1945:63 I; Nordisk ministerråd Rapport 1992:604).109 Concerning other tenure and investor categories, policy goals became more explicitly formulated as of the 1970s. In Norway, governments regarded it as “desirable that as many as possible own their dwelling, either by themselves or through a cooperative” (“samvirke”). This aim was supplemented by an “acceptance” of rented housing for economically weak persons at the end of the 1980s 127 (St.meld.nr. 76 1971-2; St. meld.nr. 34 1988-89; St.meld.nr. 23 2003- 2004). In Sweden, the official housing policy was to be “tenure neutral” since the 1970s (Prop. 1974:150; Nordisk ministerråd 2001). However, such general formulations served only as rough guidelines for adopted policies, which specified and varied such broad and more consensual formulations. The adopted housing policies included a wide range of measures, such as municipal land control, tenant protection, housing allowances and capital subventions (see Appendix C for an overview of policy interventions in housing). Municipalities’ right to purchase land for site leasehold was introduced in 1907 (Sweden) and in 1916 in Norway (ILO 1930). Afterwards, municipalities gained increasing rights to expropriate and plan mandatory use of land. The strongest regulations for municipal land control were introduced in 1947 in Sweden (delimited in 1993) and in 1966 in Norway (Bostadsstyrelsen 1968; Boverket 1994). In Norway, land allocation officially discriminated in favour of cooperatives. For example, municipalities were officially advised to work with cooperatives when preparing and planning new construction sites (Lov om den Norske Stats Hus Bank 1946). The cooperation between municipalities and cooperatives increased in 1964, when “construction site companies” were founded in order to purchase, plan and preparate construction sites; these were owned in thirds by municipalities, regional cooperatives and the national cooperative umbrella organisation. In 1982, municipalities’ positive discrimination in favour of cooperatives was abolished (Reiersen and Thue 1996). In Sweden, municipal land allocation was officially non- discriminative between investors. However, municipal influence on land allocation has often been claimed to favour cooperative and public investors while putting private investors at a disadvantage (Lundqvist 1988). Rent level controls, which are often regarded as a strong disadvantage to private landlords (McCrone and Stephens 1995, Gulbrandsen 1988, BoU 1998/9:3), were established 1939 in Norway and 1946 in Sweden.110 In the 1970s, Sweden’s rent control system was turned into a collective bargaining system, under which rents were negotiated by the National Tenant Association and municipal public utility companies, which in turn became binding for all types of landlords. In Norway, rent controls were replaced by municipal rent boards, which act as arbitrage institutions for tenants and landlords.

128 A further type of statutory intervention was the provision of housing allowances, usually directed towards specific types of households. In Norway, housing allowances are only granted to households of private landlords if children belong to those households, while households with other tenure forms are eligible irrespective of family responsibilities (after means testing and if they are receiving social benefits as families or pensioners). In Sweden, housing allowances are granted irrespective of tenure or ownership status of families and pensioners after means testing (NOU 2002:2). Swedish and Norwegian housing policy allows generous tax reductions for interest rate payments (as do most other Western countries). Deductibility of interest payments from income tax is a costly subvention in the form of foregone tax revenues in both countries and is usually regarded as the single most expensive housing policy. In the case of rental tenants, the tax reductions are claimed by the landlord, while home-owners are allowed to deduct interest payments individually from their incomes. Likewise, cooperative owners can deduct interest for the share of the capital they financed, but for the major share of investment capital, the deduction right lies • with the cooperative. In Norway, tax deductions for cooperatives were introduced in 1976, but are linked to the owner’s marginal tax rate and have an upper limit. In Sweden, tax deductibility was delimited to 55% in 1982 and to 30% in 1991. According to Lundqvist, owner- occupation is the most strongly subsidised tenure at the household level, mainly due to the possibility to deduct interest payments form individual income taxes (Lundqvist 1988, 1992c). Real estate taxation eventually also discriminated between investor categories. In Sweden from the 1940s until the 1990s, private multi- dwelling ownership was taxed in the same way as other properties, while home, municipal and cooperative estates were charged specially calculated taxes, which amounted to a lesser tax burden. After 1993, however, real estate taxation was equalised for all owner categories, with the exception of cooperative property, which continued to be taxed under the old system (for further discussion of tax differentiation between investors and ownership types, see Lundqvist 1992a, c and Boverket 1994). In Norway, dwellings built with the help of public loans were in principle free of real estate taxation during the first 20 years after their establishment. This implied an advantage for those investor categories that obtained public loans, mainly home-owners and cooperatives.

129 Furthermore, real estate taxation is generous in that assessed values of dwellings increase slowly. Besides policies that indirectly affected investor and tenure categories, eventually tenure and tenure conversion regulations were also applied. Due to the special construction of Swedish estate property laws, ownership of single dwellings in multi-family houses (condominiums) is unfeasible (Victorin 1987). This can be interpreted as an advantage for all other tenure types in multi-dwelling houses, but especially for cooperatives, which offer a comparable tenure. In Norway, condominiums were in principle legal, but the number of projects to build new condominiums remained limited until the 1970s (Gulbrandsen 1988). Regulations concerning conversions from one tenure type to another became more important since the 1970s in both countries. In Sweden, conversions from the large municipal housing stock were increasingly allowed, favouring private landlords (Boverket 2005), while in Norway, conversions from the cooperative housing stock were gradually liberalised, favouring predominantly condominiums. In sum, the above-mentioned Swedish and Norwegian housing policies to some degree differentiated between investor categories or tenure types. It could be argued that from the 1940s to the 1980s, private landlords were discriminated against in various ways, for example through rent control, tenant protection, real estate taxation and, partly, land allocation and housing allowances. On the other side, individual home-owner investors have been favourably treated as a tenure category in Norway and, to a lesser degree, in Sweden, since they can deduct interest payments from income tax in combination with other subsidies. Public and cooperative investors fall in between these categories due to policies such as favourable real estate taxation, housing allowances and land allocation. After the 1980s policies tended to became less discriminative between investor and tenure categories, with the exception of home-ownership. In the following, I will focus on policies of capital subvention and related conditions for investors and owners of housing. Among the various policy measures adopted, capital subventions are often regarded as the “cornerstone” of public housing involvement, especially during the decades of high construction activity and housing subvention from the 1940s to the 1980s (Bengtsson 1992). Likewise, Lundqvist regards that “public involvement in production

130 finance set the stage for the rest of the public-private mix” (Lundqvist 1992a).

Sweden Capital subventions came in various forms, such as loans, interest rate subsidies or a combination of both. In the early years, even simple lump-sum subsidies were used. Statutory loans and subsidies were issued through public banks and directed by the National Housing Board (Bostadsstyrelsen, since 1988 Boverket). Loans were publicly financed until 1986, when they moved to open markets through the newly founded State house Company (SBAB). Statutory intervention in the housing market can be divided into three main phases concerning the level and coverage of subsidies for new construction activity. During a phase of low-level intervention until the 1940s, time-limited measures were issued to interfere in the housing market, and the share of subsidised new dwellings remained limited, ranging from 7% to 45% (see also previous section). From the 1940s onwards, a high intervention phase set in, lasting • until the beginning of the 1990s. During the decades of high housing subvention, a “permanent” subsidy system was in place, providing a stable subsidy frame for housing investors. The coverage level increased to around 80% in the 1940s and has since remained at around 90-100% of all new dwellings. The high share of subsidised dwellings since the 1940s implies that only a few housing projects have not been influenced by conditions set through statutory capital subventions. The share of subsidised dwellings remained high even in the 1990s, but the “depth” of subvention has decreased since the early 1990s (Boverket 2005). In several steps, guaranteed interest rates were raised, subsidies to cover development costs for construction activity were lowered and several housing-related taxes were raised, such as real estate taxes and value-added taxes (VAT). The level of intervention from the 1940s to the 90s is usually regarded as “ambitious” (Heady 1978, Balchin 1996a), while the changes at the beginning of the 1990s are regarded as “deregulation” (Turner 2001) or even as the “end of an epoch” (Strömberg 2001).111 How were conditions shaped for different investor types? Until the 1940s, all investor types were in principle eligible for the same conditions. However, subvention policies were formulated with

131 preferences for public utility investors (e.g., loans in the 1920s, SFS 1922:413, SFS 1929: 127). This changed during the main intervention phase from the 1940s to the beginning of the 1990s, when lending conditions were differentiated according to investor types. During this period, statutory capital subvention policy was varied, first, concerning a guaranteed fixed interest rate on statutory second mortgage loans (eventually, also on first mortgage loans from private banks), and, second, concerning the share of development costs covered by statutory loans, which, in general, were second mortgage loans above 70% of development costs. This meant that the state took on riskier loans, but investors first had to borrow mortgages from private banks. First, the level of guaranteed interest rates was more advantageous for all categories of multi-dwelling investors than for private home- owner investors.112 Second, loans covered to various degrees development costs of housing projects, which were “approved construction costs,” ranging from 85 to 100% of development costs. A third distinction was made concerning the size of statutory subsidised loans. Statutory loans covered between 15 and 30% of the approved development costs. Thus, the loans at the highest risk level, which had to be provided by investors themselves or to be loaned on the private market without guaranteed interest rates, varied between 15 to 0% . In both respects, (municipal) public utility investors were treated most favourably. They could obtain loans calculated on 100% of their development costs, including a statutory loan covering 30% of their development costs. Cooperative investors had nearly the same conditions: they were eligible for loans based on up to 95-99% of their costs and a direct loan of 25-29% (the remaining capital had to be covered by the individual owners of cooperative dwellings).113 Finally, private rental investors were least subsidised. Their cost coverage went only up to 85-95% with loans of 15-25%. In 1991, the level of housing subvention was reduced. However, the subsidy system with its differentiation between investor categories remained in force.114 This changed in 1993, when the old subsidy system was replaced by a new one. “Approved development costs” under the new system were no longer varied between investor categories, but all types of investors became eligible for the same coverage levels, meaning that

132 municipal and cooperative investors were no longer treated favourably in comparison to other investors, including private landlords (Boverket 1994, 2005).115 In sum, policy programs differentiated between investors most clearly during the period 1942-1991. Policy conditions during this time were favourable for public and cooperative investors, giving them a competitive advantage over private rental property investors.

Eligibility regulations for associations Eligibility for public subvention is an interesting financial aspect during the period between 1942 and 1992, due to the high level of public involvement. As has been described above, favourable conditions were given to cooperatives and, especially, public utility companies. Which types of companies were acknowledged as public utility companies? Public utility companies were originally founded – often co- founded – by associations and municipalities. Information about the social background of those housing associations is scarce. In official statistics, specification is mainly given in the form of the public utility • status of such companies, but, eventually, associations are characterised as philanthropic organisations, foundations, associations or as belonging to the Swedish State Church. During the 1930s and 40s, even cooperatives were involved in public utility companies in connection with special housing programs for families and pensioners. It is unclear which housing among that identified as employers’ housing, which was acknowledged as “public utility” during the 1920s, was still organised in public utility companies in the 1930 and 40s (SOS Population and Housing Census 1920, 1933, 1945, SOS Byggnadsverksamheten various years). Since 1939, to be publicly acknowledged as public utility companies organisations had to operate as limited profit companies, out of which private capital could only be extracted on its nominal value at the time of investment. In addition, for an organisation to qualify as a public utility company “more than the majority of the housing board” had to be municipally appointed in 1948 (SFS 1948: 587-588). Furthermore, from 1986, the entire board of a public utility company had to be appointed by the municipality, and all of the capital had to originate from municipalities (SFS 1986:694).

133 In other words, associations organised into public utility companies lost their independence and, thus, their distinction from public administration in 1948. Conditions for public acknowledgement were changed again in the 1990s, at which point associations could again be recognised as self- administered public utility investors; that is, acknowledgement as a “public utility” no longer required submission to municipal control and capital (SFS 2002:102). However, today such recognition no longer has the same implications concerning public subvention because all investors receive subsidies under the same conditions. In sum, statutory support for associational housing before 1948 was provided for a broader range of associations, including cooperatives and others. The legal change in 1948 of requirements for public utility acknowledgement implied that, afterwards, cooperatives were left as the only independent associational investors eligible for more favourable treatment than private investors. As a result, cooperatives had better conditions than other associational investors during the period of high public intervention from the 1940s to the 90s (see table 20).

Norway In Norway as in Sweden, the level of statutory subvention for construction varied. The Norwegian central state did not concern itself very much with active housing policy up to the Second World War.116 This changed in 1945, when the central state began to intervene in the housing market through funding housing investment on a larger scale (for other policies, see above). Issuing loans, and monitoring and advising investors were delegated to the newly established Norwegian State Housing Bank (Den Norske Stats Husbank) and the Housing Directorate (Boligdirektoratet). These were controlled by the government through regulations and budgeting. Unlike in Sweden, public loans usually cover the first (and second) mortgages. From 1946 until 1982, Norwegian State Housing Bank loans could in principle cover up to 100% of the construction costs, but usually it covered around 75% (see Appendix A table A31 for regulations).117 The interest rates of public loans were subsidised and repayment regulations were generous. As a result of the new housing policy, 66% of all initiated housing developments were built with the help of public loans from 1946 to 1982. The development costs of single

134 construction projects were on average 70% covered by public loans (NOU 2002:2). Since the beginning of the 1980s, the housing intervention of the central state gradually decreased. In 1982, public loans no longer covered more than 60% of the development costs and the share of projects covered was reduced. The share of completed dwellings that had been financed with the help of public loans fell from 66% to 58% from 1983 to 2004. In addition, a smaller portion of development costs of single housing projects was covered, dropping from 70% to 63% from the 1980s up to 2004 (State Housing Bank, SSB). Furthermore, subvention of interest rates for public loans was reduced in the 1990s and abolished in 1996 (interest rates of public loans became set according to market rate levels). All types of investors have been eligible for public loans since 1946. However, official recommendations to the State Housing Bank included some differentiation among investor categories. It was recommended that up to 75% of developing costs be covered by public loans for all investors, but cooperatives could receive a higher • cost coverage. De facto, both cooperatives and other investors received less, with cooperatives receiving about 7% higher cost coverage. Furthermore, it was recommended that collective investors and private persons, not private landlords, should be given preference with respect to issuance of public loans (Ot.prp.nr.12 1945-46). Recommendations were changed in 1982, when public social renting was included for the purpose of solving housing problems for specific groups (e.g., the elderly, younger persons and immigrants; St.meld.nr.34 1988-89; Annaniassen 1996c). The actual distribution of State Housing Bank loans can be seen in table 19. Loans were given mainly for developing owner-occupied and cooperative housing, while investors in rented housing, such as public authorities, received a minority of loans.118

135 Table 19: State Housing Bank loans, investor categories, shares, %

2 Cooperative Associational Public Private Total authorities1 1968-1981 32 8 0 60 100

1982-2004 17 6 7 65 100

All years 23 7 7 63 100 1 public investors received additional loans for special housing (care facilities, orphanages, etc.) 2 private persons and other private investors, not differentiated in statistics Source: State Housing Bank

Some differences can be seen between the first phase until 1980s and a second phase since then. Cooperatives received a greater share of public loans until 1981 than in the decades afterwards. Cooperatives received 32% of all State Housing Bank loans until the 1980s, a share that was two to three times higher than their housing stock shares of around 10%, while afterwards they only received 17% of all loans compared to their housing stock share of about 16%. For other investors, the reception rate developed in the opposite direction since the 1980s.

Summary Swedish and Norwegian politics subsidised and regulated cooperative housing for several decades. From the 1940s until the 1980s/90s, cooperatives were given preference over private landlords and other associations with respect to the granting of statutory loans and subsidies. In Norway, cooperatives held additional advantage in comparison to public investors (see tables 20 and 21).

136 •

Table 20: Sweden: Subventions and Associational provision, shares -1940 1948- 1960- 1980- 1990- Statutory Low High Middle Intervention level

Subsidy regulations Indifferent/ 1. Public utility companies (publicly controlled) 100% Indifferent concerning investor types Stating preferences for 2. Cooperatives 95% (multi-dwelling houses) public utility companies 3. Private, other associations 85%

Cooperative Ownership, 4 11 16 17 Housing Stock Shares, % Cooperative Tenure, 4 9 14 15 17 Housing Stock Shares, %

137

Table 21: Norway: Subventions and Associational provision, shares in % -1940 1940- 1980- 2001 Statutory Low High Middle intervention

Subsidy distribution: 1. Cooperative 31% 1. Cooperative 17% ranking by investor type, 2. Foundations, etc. 8% 2. Public 7% (multi-dwelling houses) 3. Public 0% 3. Foundations, etc. 6%

Cooperatives, 6 19 14 Shares in housing stock, %

It is argued here that the support given to cooperatives was an important factor behind cooperatives’ high levels of construction activity and expansion of housing stock shares until the 1970s in Sweden and 1980s in Norway (see also Bengtsson 1992, Lundqvist 1992a, Annaniassen 1996a, Strömberg 1992, see figures 10 and 11). Statutory support is regarded as relevant due to the relatively generous subvention of new construction activities during these decades. Furthermore, financial markets and mortgage banking were highly regulated until the 1980s, which may have increased the relative relevance of mortgage lending through the state. Municipal site planning in Norway was another factor of discrimination in favour of cooperative housing projects because municipalities were advised to work in close collaboration with cooperatives. However, favourable public policies for cooperatives cannot be regarded as the only factor explaining the relative growth of cooperatives. Another factor is that cooperatives proved able to act as reliable housing investors, with only minor cases of insolvency and, most basically, their relative popularity as providers of multi-dwelling, at-cost housing during the decades of high urbanisation and relatively low wealth in both countries (see also the next section for a discussion of the “character” of cooperative housing during these decades). A lessening of cooperatives’ popularity may underlie their already slowed growth in comparison with owner-occupied housing during the 1970s and 80s in Sweden, when policy conditions were still favourable for cooperative housing. Shifting (middle class) household demands for owner-occupied detached housing may explain the relatively slowed success of cooperatives (see Appendix A figures A1 and A2 for construction of detached houses). Since the 1980s and 90s, housing policies changed towards cooperatives as investors as well as towards the cooperative tenure. First, public capital subvention was readjusted, whereby cooperatives became equalised with other investor categories in Sweden and cooperatives’ share decreased in favour of public investors in Norway. In Norway, additionally, statutory regulation of municipal site planning shifted from recommending close collaboration with cooperatives to a more neutral consideration of all investor categories. Second, implications of public housing policies for cooperatives can be seen in regulations that allow greater freedom concerning conversions of tenure and ownership structures since the 1980s in

138 Norway and since the 1990s in Sweden (for policy regulations, see sections 4 and 5). In Sweden, conversions from public rented to cooperative tenure have increased and are regarded as the most influential factor for the growth of cooperative tenure since the 1990s (Bengtsson 2006). In Norway, on the other hand, cooperative housing itself was more often being converted into condominium housing, which is regarded as one reason for the diminished share of cooperative housing (Hansen 2003). Changed public conditions for cooperatives may have been one reason for the more unstable development of cooperatives since the 1980ies. In Sweden, overall cooperative ownership and tenure still increased, but cooperatives’ construction activity became more volatile in comparison with other multi-dwelling housing investor categories until the 1990s; that is, cooperatives lost their former position as constant “second” active housing investors, but alternated in that position with private investors. In Norway, cooperative construction activity decreased in comparison with other investor categories and, likewise, cooperative • tenure and ownership decreased in comparison with other housing forms since the 1980s. In sum, it is here argued that statutory housing policies positively influenced – although they did not determine – the relative success of cooperatives during the first decades after the Second World War. Furthermore, it is argued that housing policies have become more mixed concerning cooperative investment and cooperative tenure, influencing the more mixed development of cooperative investment and tenure since the 1980s and 1990s. How political advocacy and opposition changed over the decades in Sweden and Norway is the theme of a later section. First, the state regulations concerning cooperative housing and the involved cooperatives will be described in greater detail.

139 80% y, % y,

70%

60%

50%

40%

30%

20%

10%

0% share of completed dwellings by investor categor investor by dwellings completed of share 1949 1954 1959 1964 1969 1974 1979 1984 1989 1994 1999 2004

public cooperatives private

Figure 10: Sweden: Completed dwellings in multi-dwelling buildings by investor category Note: * Figures for Sweden and Norway are not comparable, because they display construction of dwellings in multi-dwelling houses in Sweden and construction of dwellings in all form of houses in Norway (seeAppendix A figures A1 and A2 for completed dwellings in detached houses in Sweden). Source: SOS Byggnadsverksamheten, SOS Bostads - och byggnadsstatistisk årsbok, various years

80%

ry, % ry, 70%

60%

50%

40%

30%

20%

10%

0% 1953 1962 1971 1980 1989 1998 share of completed dwellings by investor catego dwellings investor by completed of share private public other cooperatives

Figure 11: Norway: Completed dwellings by investor category Note: * Figures for Sweden and Norway are not comparable, because they display construction of dwellings in multi-dwelling houses in Sweden and construction of dwellings in all form of houses in Norway. Source: SOS Byggnadsverksamheten, SOS Bostads - och byggnadsstatistisk årsbok, various years

140 4. The involved associations: cooperatives of differing character Housing cooperative is a term comprising a variety of organisations (see introduction housing). Here the main types of cooperatives in Sweden and Norway will be described, with an emphasis on the internal structure and operational methods of cooperatives. In that regard, statutory regulations pertaining to cooperatives, as well as self- regulation, are described.

Cooperatives of different character in Sweden and Norway Cooperatives are usually business organisations that are democratically ruled by their stakeholders. They differ from corporations in that they apply a one-member-one-vote rule, with each vote having equal weight, regardless of the number of shares held by individual members. In addition, restrictions on gains and the use of gains by cooperatives are often applied (for overviews of variations and principles of member influence in cooperatives, see Fischer 1995, International Cooperative Association-ICA). In Norwegian and Swedish housing, two main types of • cooperatives can be distinguished: those affiliated with superordinate organisations, and those that are independent. Affiliated cooperatives own around two-thirds of cooperative dwellings in both countries, while independent cooperatives own around one-third (see Appendix A table A32). Examples of superordinate national organisations are HSB Riksförbund (Hyresgästernas Spar- och Bostadsförening), Riksbyggen and NBBL (Norske Boligbyggelag Landsförbund). Superordinate regional organisations include Oslo BOS, Trondheim BB, HSB Stockholm, HSB Sydvästra Götaland, etc. In Sweden and Norway, superordinate organisations of such cooperatives are responsible for planning, coordinating and administering construction projects, while single cooperatives administer buildings after construction is completed (for a more specific description see Lundevall 1992; Annaniassen 1996a; ILO 1964; Svensson 1993). Independent cooperatives, on the other side, are not members of superordinate organisations, although some establish management or administration contracts with superordinate cooperatives. Affiliated cooperatives are also called “open” cooperatives, while independent cooperatives are usually “closed.” These terms refer to

141 the different role of membership and habitation within cooperatives. “Open” cooperatives do not restrict membership to persons occupying a dwelling, but allow it to anyone interested in joining the organisation. By opening to persons outside the circle of occupants, cooperatives gain members who are interested in further construction activity. The interests of non-occupant members are further strengthened by the possibility of saving in assimilated banks. In addition, members can arrange to queue for new or vacant dwellings. Such queuing systems are partly also used to limit market price building in giving senior members pre-emptive buying rights for dwellings. Independent or “closed” cooperatives, on the other side, are established to develop a limited number of dwellings for (incoming) occupants. (Incoming) occupants are the only members of closed cooperatives, signing up for a specific dwelling in advance. After a planned number of dwellings has been constructed, the cooperative stops developing and constructing new dwellings. Usually, such closed cooperatives are founded by construction firms or private persons (Föreningen kooperativa studier 1999, Victorin 2003). In Norway, the organisational form of “joint-stock company” is also used, which operates in a similar way as independent cooperatives. Cooperative ownership can come with free-market disposal methods similar to ownership of condominiums, or with price and sale limitations on individual ownership titles, more similar to social- collective housing (see introduction housing). In addition, the initiating company may either sell dwellings at cost to new owners or seek to sell at a profit. Independent cooperatives tended to be of a more private-market character in Sweden, while, in Norway, independent cooperatives or joint-stock companies had a more mixed character, because they were often subjected to similar statutory regulations as cooperatives. Affiliated cooperatives, on the other side, were originally of a more social-collective character by being affiliated with a superordinate cooperative that attempted to delimit dwelling prices. In Norway, affiliated cooperatives offered new dwellings at cost and required that older cooperative dwellings be sold to senior members. Long-term members had a pre-emptive right to obtain dwellings before new members or non-members. Such limitations were meant to dampen prices on older cooperative dwellings (Annaniassen 1996a, b). In Sweden, affiliated cooperatives sold new dwellings at cost and

142 enforced price and access regulations so that inhabitants would sell their older dwellings at a price based on production costs (Svensson 1993, Lundevall 1996). Today, however, affiliated cooperatives’ dwellings are also sold at market prices, due to changed statutory regulations (see below for further elaboration in connection with statutory regulations). Affiliated cooperatives were introduced in 1924 (HSB) in Sweden and 1935 (OBOS) in Norway. Before the introduction of the “open” affiliated model of cooperative organisation, closed independent cooperatives were the usual form in Norway and Sweden.119 Independent cooperatives’ shares decreased after the 1940s and comprised only a minority of all cooperative dwellings since the 1960s and 70s (see Appendix A table A32). During the same decades, affiliated cooperatives increased their housing-stock share within both countries (see table 22).

Table 22: Cooperatives’ ownership, affiliation and independence, housing stock shares, %, calculated* 1945 1960 1985 1995 1997 S Affiliated cooperatives, 2 6 10 12 12 • housing stock share, % All cooperatives, 4 11 16 17 18 housing stock share, % Total 90377 299007 584414 683000 736000 N 1950 1970 1980 1990 2000 Affiliated cooperatives, 1-2 9 14 12 12 housing stock share, % All cooperatives, 6 13 19 19 14 housing stock share, % Total 168563 258570 336807 277000 *calculations based on cooperatives’ own statistics and official statistics, not comparable Source: SOS Kooperativ verksamhet; Lundevall 1985, 1995; Föreningen kooperativa studier 1995, 1999, NSD; NBBL annual reports, Annaniassen 1996b, Gulbrandsen 1988, HSB annual reports, Riksbyggen annual reports, SOS Population and Housing Censuses, SOS Bostads- och byggnadsstatistisk årsbok, SCB, SSB, NOS Population and Housing Censuses

In both countries, affiliated cooperatives are still the most common type. In recent decades, however, independent cooperatives have had a revival in Sweden and Norway. According to construction statistics, cooperatives initiated by firms and private persons increased their construction activity in comparison with affiliated cooperatives since the 1980s in Sweden and Norway (see figures 12 and 13).

143

40%

y, % y, 35%

30%

25%

20%

15%

10%

5% share of completed dwellings by investor categor investor by dwellings completed of share 0% 1965 1970 1975 1980 1985 1990 1995 2000

affiliated independent Figure 12: Sweden: Completed dwellings in multi-dwelling houses by cooperatives Source: SOS Byggnadsverksamheten, SOS Bostads - och byggnadsstatistisk årsbok, various years 35%

30% gory, % gory,

25%

20%

15%

10%

5% share of completed dwellings by investor cate investor by dwellings completed of share 0% 1953 1962 1971 1980 1989 1998

affiliated cooperative independent cooperative, joint-stock company

Figure 13: Norway: Completed dwellings by cooperatives Source: SOS Byggnadsverksamheten, SOS Bostads - och byggnadsstatistisk årsbok, various years

144 Statutory regulations concerning cooperatives Housing Cooperative Acts in both countries regulated the form of tenure offered by housing cooperatives.120 In 1930, Sweden introduced a Housing Cooperative Act, the objectives of which were to specify the rights of tenants and to “give public protection against unsound housing association enterprises” (Svensson 1993). The owner rights included, for example, the unlimited right of tenancy, renovation rights, property rights and influence of individual owners on the association as a whole. In later decades, several amendments and changes were made to the original Housing Cooperative Act, and new acts were passed in 1971 and 1991, which did not alter the tenure form, but eventually had implications for price setting and access rights (Victorin 2003). The Norwegian specification of cooperative housing came first in 1960, which largely followed the Swedish predecessor. It differed from Sweden in that two combined Housing Cooperative Acts were passed, which regulated rights and responsibilities between affiliated and superordinated cooperatives in greater detail. In addition, queuing systems for members were stipulated in the Housing Cooperative Acts • (Annaniassen 1996a). New Housing Acts were passed in 2003 (Regjering.no). Besides the main regulation via the Housing Acts, cooperative housing was regulated through further laws. In Sweden, policy regulations concerning cooperatives often tried to influence the type of cooperatives, encouraging a shift away from individual ownership towards a social conception of ownership. Selling prices for cooperative dwellings, for example, were regulated until 1969 (Heady 1978; Victorin 1987), implying that any transfer of a dwelling had to be approved by a Municipal Rent Board (hyresnämnd). Price controls for cooperative dwellings were established to ensure that dwellings would be sold at close to production costs and not after-market demand and supply prices (sec.5 of the Act on Control of Tenant-Owner Rights). In 1969, price controls were abolished after the share of vacant dwellings reached 5.4% in 1968 and actual prices for cooperatives often could not even reach the officially allowed prices. Statutory regulation has not been introduced again since 1969, although prices of cooperatives started to increase in the 1980s (Lundevall 1997, Victorin 2003)121. With the abolishment of statutory restrictions, Riksbyggen and HSB, the main superordinate cooperatives, had abolished their own

145 price restrictions, but reintroduced various models of price restrictions during the 1980s, when prices for cooperative dwellings started to increase again (Lundevall 1997, Lundqvist 1988, Victorin 1987). However, these self-imposed price restrictions were contested by court rulings and statutory regulations. Riksbyggen’s model was prohibited by statutory law in 1982, and HSB’s model was declared illegal by the Market Court. In turn, HSB’s model was legalised in 1985 (after the Market Court had declared it illegal), thus ensuring that regional sub-cooperatives were given the contractual right by their affiliated cooperatives to purchase dwellings at prices calculated on production costs within the first 3 years (Lundqvist 1988, Lundevall 1996, HSB and Riksbyggen). Both superordinate cooperatives used this legalised model for some years. However, the price limitations excluded all dwellings older than 3 years and thus the main share of cooperative dwellings on the market. HSB pressured for an extension of the regulation up to 5 years, but was not successful (Lundevall 1996). Finally, in 1992, this self- imposed price regulation was also prohibited (SFS 1992:505). Likewise in Norway, statutory regulations made price restrictions and access regulations mandatory for cooperatives, which restricted their free disposal by individual owners. First, selling prices were limited for cooperative dwellings in various forms. Municipal Rent Boards set local maximum prices for all cooperative dwellings until the 1980s. Prices for cooperative dwellings were set well below market levels until 1982 (St.meld.nr. 34 1988-89). During the 1980s, price setting was deregulated. First, prices for independent cooperatives were freed from municipal price limits in 1981 (for the different types of cooperatives, see next section). Prices for associated cooperatives would still be controlled through Municipal Rent Boards, but were set in such a way as to allow continually close relations to market prices (St.meld.nr. 61 1981-2). Actually, maximum prices for cooperative dwellings doubled at the beginning of the 1980s and exceeded market levels in many municipalities (St.meld.nr. 34 1988-89; Annaniassen 1996b). In 1988, dwellings of affiliated cooperatives older than 7 years were freed from maximum price setting, and in 1999 public price regulations were entirely abolished (Annaniassen 1996b; Regjering.no). Since 2003, investors that had helped to finance cooperative construction projects were allowed to crucially influence statute formulations, including statute vetoes for prices and sales. Such investors were predominately

146 municipalities and the State Housing Bank. The vetoes were not extinguished if cooperatives dissolved themselves (the “vedtektsveto,” Boligbyggelagslov 2003). Second, statutory regulations also limited single cooperatives’ ability to dissolve themselves as cooperatives in order to split their property into condominiums. This regulation was introduced at the request of NBBL, the national superordinate organisation of cooperatives, after several cooperatives had transformed themselves into condominiums in Oslo (Annaniassen 1996a, b). Introduced in 1974 and re-affirmed in 1978, it was, however, abolished for independent cooperatives in 1981 (Gulbrandsen 1988). To dissolve themselves, affiliated cooperatives needed the approval of the Municipal and Housing Department; such approval was rarely given (only under conditions of insolvency or other “special reasons;” §78). This limitation for affiliated cooperatives was lifted by the new Housing Cooperative Act in 2003 (Burettslagslov 2003). Third, individual access to cooperative dwellings was regulated (Reiersen and Thue 1996; Annaniassen 1996a, b). As described above, affiliated cooperatives gave members privileged access to dwellings • that were up for sale, through offering them available dwellings before offering them to other prospects. The selection of purchasers was further restricted through a queuing system, whereby members of affiliated cooperatives were offered dwellings according to their place in the queue. Thus, sellers of cooperative dwellings could not choose their buyers based on who would pay the most. These regulations were in turn re-affirmed by the Housing Cooperative Acts of 1960, which obliged all affiliated cooperatives to provide such a queuing system. However, since the Housing Cooperative Acts of 2003, superordinate cooperatives have no longer been allowed to provide queuing systems for wider regional areas; only single cooperatives are still free to apply queuing systems for their (delimited number of) members. Furthermore, single cooperatives can themselves decide which persons or categories of persons are eligible for pre-emptive buying rights for dwellings. Fourth, the Housing Cooperative Act of 1960 had originally assured the right of cooperatives to create a statute of permanent affiliation between single cooperatives and superordinate cooperatives. Since 2003, binding affiliation is no longer allowed for more than ten years, after which single, affiliated cooperatives have to decide whether to prolong the association.

147 Fifth, statutory regulations started to infringe upon cooperative tenure and organisation by providing “municipalities, employers and others” the right to own up to 30% of the dwellings within single cooperatives in 2003, even against the will of cooperatives (this right was first stipulated in 1995, when a 10% ownership was allowed for; Boligbyggelagslov 1990, 2003).122

Summary Affiliated cooperatives became the predominant form of cooperative, while independent cooperatives became the minority in Sweden and Norway. Affiliated cooperatives’ shares of the housing market rose constantly after the Second World War and reached their highest levels in 1980 in Norway and the 1990 in Sweden. The relative success of affiliated cooperatives, in comparison with independent cooperatives, may lie in their ability to provide low-cost housing for members of the middle classes. Originally, a main difference between affiliated and independent cooperatives was that affiliated and superordinate cooperatives mandated crucial aspects of access and price formation for their members, which made them social-collective housing organisations, while independent cooperatives operated more according to market incentives, which made them more like condominium housing providers (see also introduction for housing). The relative success of affiliated cooperatives in comparison with independent cooperatives over several decades may have been grounded in their ability to efficiently and effectively build and administer housing and to prevent large-scale insolvencies for members of the lower to middle class strata, while independent cooperatives and joint-stock companies were focused on producing more expensive dwellings for the relatively smaller social strata of the upper-middle classes. Affiliated cooperatives could even sell part of their relatively well-developed managerial and administrative capabilities to independent cooperatives, indicating their relative organisational superiority. Nevertheless, some changes have occurred in recent decades in both countries, which could call into question the former success of the affiliated type of cooperative. First, independent cooperatives increased their construction activity in comparison with affiliated cooperatives since the 1980s. Second, affiliation lost its original social-collective distinctiveness. Although superordinate cooperatives

148 still sell newly built dwellings at production costs and offer them to members according to their place in a queuing system, affiliation has far fewer implications for gaining access to and selling old cooperative dwellings, which comprise the largest share of cooperative housing stock. In this regard, superordinate cooperatives developed into mere service organisations for their affiliated members and, in turn, the situation for single, affiliated cooperatives is more like that of independent cooperatives or condominium organisations. It is difficult to assess the extent to which statutory interventions have influenced the relative success of affiliated cooperatives in the housing market. Statutory construction subsidies distinguished between the different types of cooperatives to some degree in Norway, where independent cooperatives received subventions for far smaller shares of their construction projects than affiliated cooperatives (see Appendix A table A33). On the other side, such differentiation did not occur in Sweden, thus not accounting for their relative success in Sweden. Likewise, municipal construction site planning could have been a reason for affiliated cooperatives’ success in Norway, where, by law, • they were given preferential consideration until 1982, but again not in Sweden (where, however, municipalities could have given de facto advantage to affiliated cooperatives, but not due legal prescription). Another possibility is that statutory policies set “matching” incentives for affiliated cooperatives with further regulations. States provided subsidies for construction projects, which were approved as being within certain minimum and maximum standard levels since the 1940s and 50s, for example, with respect to number of rooms, floor space and equipment (Lov om Den Norske Stats Husbank 1946, Rhune and Theiersen, SOU 1945:63 I, Nordisk ministerråd Rapport 1992:604). Such standards may have been a disincentive for independent cooperatives, which traditionally had built for an upper- middle-class clientele. Affiliated cooperatives, on the other side, had traditionally built for lower-middle classes and were, thus, not restricted by such standards. Furthermore, affiliated cooperatives had accumulated particular experience and expertise for such lower-cost construction, including architectural knowledge and contacts with suppliers and construction firms. Thus, they may have been better prepared to combine the incentives set by policies and the demands of rising working and middle classes. Subsidy retrenchments since the 1980s in Norway and the 1990s in Sweden may be a reason for the

149 decrease in affiliated cooperatives’ construction, due to difficulty adapting to new groups of buyers and new financing conditions. In addition, price restrictions on cooperative dwellings, which were in force until 1969 and 1981 respectively, were binding for independent as well as affiliated cooperatives. On one side, it cannot be assumed that affiliated cooperatives were discouraged by such price regulations, because they had themselves decreed price regulations and at-cost construction in by-laws and, in addition, had often urged states to decree sales restrictions. Independent cooperatives, on the other side, may have had more interest in producing for better-off clients, for whom obtaining lower-priced real estate was not a preference.123 The relevance of statutory regulations for affiliated cooperatives is obvious when it comes to regulations concerning their operation. Until the 1980s, the Swedish and Norwegian states tended to support “social-collective” statutes of cooperatives, either through permitting cooperatives to independently set their own regulations or through incorporating cooperative statutes into statutory regulations. Examples of legal incorporations are price regulations and the Norwegian Housing Cooperative Acts of 1960, which explicitly allowed for permanent affiliation and mandated member queuing systems resembling those used by cooperatives in earlier decades (Lundevall 1994, 1996, Annaniassen 1996a, b). Since the 1980s and 90s in Sweden and Norway, however, statutory regulations may, in turn, have been less favourable for affiliated cooperatives. Statutory laws tended to move cooperative housing towards individual ownership by strengthening free-market access and disposal rights, but weakening social and collective ownership and cooperatives’ self-regulation. This happened, for example, when dissolution rights were granted to affiliated cooperatives in Norway, and when self-imposed price regulations were made illegal in Sweden. A second trend of infringement on cooperative self-regulation can be observed in Norway: “Municipalities, employers and others” have the right to acquire cooperative dwellings against the will of cooperatives and can influence the statutes to impose price and sales restrictions. In the next section, the focus will be on political parties’ influence on state policies. The question is whether party policies diverged concerning their preferences for statutory regulation of cooperative housing. Which parties supported or opposed favourable terms of

150 financing cooperatives? And which parties supported or opposed a social, collectively regulated operation of cooperatives?

5. Partisan policies In general, political parties in Sweden and Norway agreed upon state intervention in housing. Main goal formulations for housing policies, for example, were maintained over several decades, irrespective of changing party compositions in government. Examples of goal formulations for the post-war decades until the 1980s and 90s were, in Sweden, that “all should be able to dispose over a good dwelling at costs that are in reasonable proportion to incomes” and, in Norway, that “healthy, well-planned and practically equipped dwellings of good quality should be prepared for the whole population at reasonable costs” (SOU 1945:63 I, Lov om Den Norske Stats Husbank 1946; Prop. 1967:100; St.meld.nr. 76 1971-2). Since the 1980s, goal formulations eventually became less ambitious and increasingly included environmental aims. In Sweden, housing policies are today meant to ensure that “all should dispose over a good • dwellings in good housing environments” and, in Norway, that “all should be given the means to live in good dwellings at a reasonable cost and in a stimulating and safe, ecologically sustainable environment ” (NOU 2002:2). Such general policy guidelines were accompanied by details such as official definitions of “overcrowding” and housing expenditure/income ratios of 20-25%, which were used as yardsticks for the overall housing situation up to the 1990s (Lov om Den Norske Stats Husbank 1946, Rhune and Theiersen, SOU 1945:63 I, Nordisk ministerråd Rapport 1992:604). Likewise, there was political agreement about the general level of statutory intervention in the housing market. The beginning of high intervention in housing policy in the 1940s was voted for by governing and opposing parties in both countries and maintained by both right-wing and left-wing governments in the following decades until the 1980s. In practice, right-wing and left-wing governments allowed house construction and public subsidies to grow until the 1980s.124 From the 1980s onwards, governments of different political party compositions attempted to lower public housing expenditures. After these changes, the intervention levels remained limited in both countries.125

151 While there was political agreement over the general outlines of housing policies, there were controversies centred around more detailed questions of subsidies, taxes and regulations, which in turn affected relative support for different household categories, and, of interest here, investor and tenure categories (for an overview of housing policies, see section 3). In order to describe party standpoints in conflicts about policies relevant for associational housing, the time span since the Second World War is dealt with in two phases: a first phase until the late 1980s in Sweden and the beginning of the 1980s in Norway, and a second phase since then, respectively.

Sweden Phase I: 1945-1990 From the 1940s to the late 1980s, controversial political decisions that have been important for the relative prevalence of cooperatives concerned three policy areas: first, the level of subventions for associational providers in comparison with private investors; second, eligibility regulations for associational investors; and third, cooperatives' self-regulated operation. The relative subvention level for associational providers was changed two times: first, in the 1940s and, second, in the 1990s. In 1946, cooperatives became eligible for higher subvention and loan levels than private investors, but lower levels than public utility companies (allmännyttiga bostadsbolag). A social democratic government initiated and enforced this regulation against the criticism of the opposition parties. The Conservative and the Liberal Parties, especially, demanded equal subsidy levels for cooperatives and private investors. Both parties voted against a more favourable subvention of cooperatives under the new housing policy scheme (although they supported the new policy scheme in many other aspects; Prt. II 1946:30). They argued that cooperatives should be understood as a private form of housing. The Social Democratic Party admitted that cooperative housing had the flaw of demanding a capital investment in order to have access to a dwelling. Already in 1945, the social democratic-led housing commission stated that the demand for cooperative housing had reached a surfeit because lower-income groups were not capable of making the necessary advance payment. Therefore, they regarded

152 rented housing as the most feasible form of housing to provide for broader social strata (SOU 1945: 63 I). However, based on experience from the years before the Second World War, the Social Democratic Party still saw cooperative housing companies as reliable partners for the development of non-speculative at-cost housing, which would contribute to an enhanced housing supply and housing standards (SOU 1945: 63 I). The subsidy scheme of 1946, with more favourable conditions for cooperatives than for private investors, was upheld until 1992, when a right-wing coalition displaced it with a new system. Then, a coalition of the Conservative, Liberal and Centre Parties introduced a subsidy system providing equal subsidy levels for all investor categories. In so doing, the right-wing coalition also abolished the relatively advantageous subsidisation of cooperatives (and public utility companies). The intention of the new subsidy system was to increase the “interest rate sensitivity” of investors, to “abolish concomitant shortcomings concerning the neutrality between different tenure forms” and to “liberalise the housing market” (SOU 1992:47, cited in Boverket 1995, Strömberg 2001; Prt.1991/2:30, cited in Persson • 2001). The Social Democratic Party opposed the introduction of the new scheme, but did not change it after regaining power. Instead, they provided additional special short-term subsidies for rented housing. Further controversy in the 1940s arose around the conditions for the acknowledgement of providers as public utility companies, which received the highest subsidy levels from 1946 to 1992. Public utility companies were run by either municipalities, associations or both since the beginning of the century. In 1948, this liberal praxis of recognising differently composed companies was changed. A precondition for being acknowledged as a “public utility” – and, thus, for receiving the highest subsidy levels under the subsidy system of 1946 – became that municipalities appointed the majority of board members of such companies. The shift towards tighter municipal control was enforced by a social democratic government, but was met by critics in parliament. The Conservative, Liberal and Centre Parties voted against the introduction of municipal control. They argued that municipal control unduly excludes “committed persons” from public utility housing (Prt. II 1948:27). This change left cooperatives as the only independent associational investors eligible for higher subsidies than private investors.

153 Different conceptions of the Social Democratic Party on one side and right-wing parties on the other side appeared also concerning cooperatives’ rights to self-regulate their operation. National cooperatives Riksbyggen and HSB (re-)introduced price regulations for dwellings of their members in the 1980s. Both cooperatives introduced pre-emptive buying rights of cooperatives or municipalities for dwellings, offered by individual owners. The aim was to secure at-cost selling prices. A right-wing coalition (the Conservative, Liberal, and Centre Parties) outlawed a Riksbyggen-scheme in 1982 and, thus, left the right of decision to the individual owner (Prop. 1981/2:169). In addition, the coalition remained passive when the Market Court declared an HSB regulation illegal. By contrast, the subsequent Social Democratic government legalised the HSB scheme of self-regulation (BoU 1984/5:12). Cooperatives gained the right to impose a regulation giving regional cooperatives the option to purchase dwellings at a close-to-cost price within the first three years (see above). After regaining power, the right-wing coalition abolished this possibility again in 1992. In sum, political differences appeared between the Social Democratic Party on the one side and right-wing parties on the other side concerning associational involvement in housing. They differed on the relative privileges of associations concerning public subsidies, the type of associations gaining such privileges and the rights of associations to regulate themselves. The Social Democratic Party in all aspects supported associations in form of affiliated cooperatives. In contrast, right-wing parties supported private investors or other types of housing associations. Furthermore, they eventually restricted associational self-regulation (see table for summary of partisan policies).

Phase II: 1990- From the 1990s onwards, political parties agreed upon changes concerning cooperatives as a formerly privileged investor category, and concerning cooperatives’ rights to self-regulate their operation. The new housing subsidy scheme of 1992 implied for cooperatives the loss of their advantaged position as an investor category in comparison with private investors. Furthermore, national cooperatives’ price restrictions for newer dwellings were made illegal

154 in 1992. Both the new subsidy scheme and prohibition of cooperatives’ price restrictions were enforced by a right-wing alliance of the Conservative, Liberal, Christian Democrat and Centre Parties. These changes were met by strong protests of left-wing parties, while in opposition, but the Social Democratic Party did not change the new subsidy system for housing when they regained power. Neither did they reintroduce cooperatives’ right to restrict prices. Further, the Social Democratic Party disregarded cooperatives when issuing new subsidy measures, but supported instead (low-cost) rented housing alone (new schemes were targeted at rented housing for special groups such as students and pensioners; Boverket 2005). Controversies between political parties did arise concerning strategies for the conversion of the large municipal housing stock and the introduction of new tenures. Right-wing parties, on one side, supported the existing cooperative tenure as an option for the conversion of the municipal housing stock, introducing regulations to make conversion into cooperatives easier by lowering the required majorities of sitting tenants to approve transformation of their dwellings. By contrast, the Social Democratic • Party raised required majorities among tenants for such conversions and introduced time-restrictions for transformations (BoU 1998/9:3; SOU 2000:95). Furthermore, political parties pursued diverging strategies concerning new tenure forms. The Social Democrats legalised a new tenure of cooperative housing, which can be described as cooperative tenantship (kooperativ hyresrätt; SFS 2002:93126). This new form was especially designed for founding cooperatives within the existing municipal housing stock.127 Members of the cooperative can, under the new form, choose either to purchase their dwelling as part of the cooperative, or to join as tenants. For purchasers, the crucial difference to cooperatives under tenant-ownership is that capital investments are not sellable at market prices, but have to be returned to the cooperative at the amount originally invested. This regulation is intended to inhibit price changes for dwellings. Gains (or losses) due to market changes are not possible for the individual occupant (SOU 2000:95). Right-wing parties, on the other side, attempted to legalise the tenure form of condominiums, that is, the ownership of single dwellings within multi-dwelling houses, which was not possible in

155 Sweden. In 1994, they set up an inquiry into owner-occupation in dwellings, but the new Social Democratic government in the same year ended it. Afterwards the conservatives regularly requested the introduction of owner-occupation for dwellings (e.g., Mot. 2002/3:214, Mot. 2004/5:285). To sum up political parties’ policies since the 1990s, it can be said that right-wing parties have partly changed their policy towards cooperatives (see table 23 for summary of partisan policies). On the one side, they have not been willing to give cooperative housing special recognition under the new subvention system, which they introduced, or to allow cooperatives to self-regulate the selling of dwellings in a social-collective way. Furthermore, the bourgeois coalition showed an interest in introducing condominiums as a new tenure type, which would presumably weaken the position of cooperatives on the housing market by reducing their relative attractiveness as a tenure for members of the upper-middle class. On the other side, right-wing parties supported cooperatives as a solution for the transformation of the municipal housing stock.

Table 23: Sweden: Associational housing and partisan advocacy

State regulations Political parties1 advocacy Subsidy generosity ranking, Associations Cooperatives multi-dwelling houses 1945/8 1. Public, other association S 2. Coop 3. Private 1948 1. Public Associations S 2. Coop publicly 3. Private controlled

1982 Coop- M, FP, C regulation of sales abolished 1985 Coop- S regulation of sales allowed 1992/3 Public=Coop=Private= Associations Coop- M, FP, C, KD other association self-controlled regulation of sales abolished 1Not included: Left Communist Party, Green Party, other parties SAP (Social Democratic Party), M (Conservative Party), FP (Liberal Party), C (Centre Party), KD (Christian Democratic Party)

156 The Social Democratic Party, likewise, changed their policy towards cooperatives. First, they did not change reduced rights of cooperatives for self-regulation, nor the new subsidy system for housing introduced by the bourgeois coalition, but, instead of re-installing cooperatives as a favoured social-collective investor category, issued subsidies only for rented housing (see section 2).128 Furthermore, they legalised a new tenure of cooperative housing as their preferred option for the transformation of the municipal housing stock. In sum, they show a more distanced stance towards cooperatives compared with the decades after the Second World War.

Norway Phase I: 1945-1980 The post-war housing subsidy scheme was introduced by a social democratic government and supported by all parties in parliament. Furthermore, political parties across party lines agreed that cooperatives and private households should be favoured over private investors, whose main purpose was seen as constructing houses for • home-owners, and public authorities, which were precluded from the subsidy system.129 Most parties remained supportive of cooperative housing in the following decades, apart from the Conservative Party, which started to criticise the subsidy system as being too one-sided in the late 1940s and 50s. They regarded the established housing policy as leading to a “construction monopoly of cooperatives under public protection” (SF 4. April 1949). However, the Conservative Party was in a minority position among right-wing parties. The subsidy system of 1945 was upheld by right-wing coalitions as well as left-wing governments until the beginning of the 1980s. More pronounced disputes between political parties arose around the moulding of housing cooperatives’ operation. In particular, the role of superordinate cooperatives became a matter of dispute in the late 1950s. In 1959, the Housing Cooperative Act was proposed, which for the first time provided a legal specification of housing cooperatives in Norway. Politically controversial was one paragraph dealing with the relation between superordinate cooperatives on one side and single, affiliated cooperatives on the other (§ 16(5)). The Housing Cooperative Act, proposed by the Labour Party government, assured the right of cooperatives to formally establish a

157 permanent relation between superordinate and affiliated cooperatives, thus, legalising the praxis of superordinate cooperatives to establish permanent pre-emptive buying rights for members, as well as price controls on older dwellings of affiliated cooperatives. The Conservative, Liberal and Christian People’s Parties regarded this right as unnecessary “organisation bureaucracy” and “restriction of member rights” and voted against this paragraph, while the Labour Party defended it as a right to assure “community” within cooperatives as a whole (Ot.prp.nr.70 1958, see also section 4). In the 1970s, disputes developed again over individual owners’ rights versus the role of superordinate cooperatives. In 1974, in response to an appeal of NBBL, the Labour government decreed that dissolution of single cooperatives from superordinate cooperatives had to be approved by the Municipal and Housing Department from 1974. Prompting this law was the fact that several cooperatives had dissolved and become independent cooperatives or condominiums to avoid the sale and price restrictions associated with affiliation. In 1978, NBBL again appealed to the government to make the 1974 law more specific. A lawyer had discovered a loophole in the original law, which had been used by many cooperatives. To solve the problem of dissolution, the Labour government amended an extention of sale and price restriction beyond affiliated cooperatives to include independent cooperatives in 1977. The Conservative Party opposed while other right-wing parties partly supported these new regulations in 1977, which, however, changed at the beginning of the 1980s.

Phase II: 1980- Parties agreed upon the introduction of a new housing subsidy scheme in 1982. Public authorities, which had so far not received statutory subsidies, should be included as a new eligible investor category.130 Public rented housing was intended to improve the housing situation of specific groups, such as immigrants, pensioners, young persons, etc. (State Housing Bank historical statistics). Cooperatives remained a favourised investor category under the new subsidy system by being eligible to higher subvention shares for their construction projects, but the inclusion of public housing as a new “social-housing” investor category apart from cooperatives did indirectly reduce the relative share of subsidies for cooperatives.

158 Besides this overall agreement on housing subsidies, controversies between parties arose around cooperatives’ price and sales restrictions and access to construction sites. In the fall of 1981, the Labour Party lost power and the Conservative Party formed a minority government with assistance from the Centre Party and Christian People’s Party.131 When they gained power in 1981, a right-wing alliance under the leadership of the Conservative Party made several legal changes, which negatively affected cooperatives’ sales and price restrictions and access to construction sites. The right-wing alliance changed municipal site land planning so that municipalities were no longer obliged to provide land first for cooperatives’ large-scale housing construction, but, instead, were to give the same consideration to small-scale site planning for detached housing. Thus, cooperatives, which had been granted special recognition concerning municipal site land policies since 1946, lost their relatively favourable access to land. Furthermore, right-wing alliances gradually increased individual owners’ disposal rights over cooperative dwellings (Innst.S.nr. 286 • 1981-82). The aim was a “more functional housing market, which should rest on balanced free dwelling conversions” (Innst.S.nr. 286 1981-2). The measures adopted by right-wing alliances were, first, a legalisation of independent cooperatives’ conversions into condominiums in 1982; and second, gradually abolishing price regulations for cooperative housing. In 1981, they abolished price restrictions for independent housing cooperatives (i.e., non-affiliated cooperatives). Prices for associated cooperatives remained under the control of Municipal Rent Boards, but prices had to be set at market levels (St.meld.nr. 61 1981-2). In 1988, the Conservative Party proposed successfully the abolishment of price regulations for associated cooperatives’ dwellings older than 7 years (the proposal was made out of opposition; St.meld.nr. 34 1988-89). Finally, in 1999 a right-wing government comprising the Christian People’s Party, Centre Party and Liberals abolished all price restrictions (NOU 2002:2). In 2003, the governing coalition of the Christian People’s Party, Centre Party and Liberals passed a new Housing Cooperative Act. Several sections of the original Housing Cooperative Act of 1960 were changed. Changes were made concerning the role of superordinate cooperatives: permanent affiliation of single

159 cooperatives to superordinate cooperatives was prohibited, and a maximum of 10 years was established. Single affiliated cooperatives were freed to make their own decisions about dissolution from superordinate cooperatives. Collective buying rights of cooperative members affiliated to a superordinate cooperative were prohibited. Only single cooperatives were allowed to have pre-emptive buying rights for their members. Furthermore, external investors, such as municipalities and employers, gained the right to acquire up to 30% of a single cooperative’s dwellings and to include a veto in the statutes concerning prices and sales of dwellings in the new Cooperative Housing Acts in 2003 (Ot.prp.nr. 30 2002-3; Boligbyggelaglov og Burettsslagslov 2003).132 The Labour Party opposed the amendments of right-wing alliances until the middle of the 1980s. However, subsequent labour governments made no effort to bring back former policies from 1986 onwards. Instead, they supported the new Cooperative Housing Acts of 2003, which further infringed on cooperatives’ self-regulation by strengthening individual owners’ and municipal and other investors’ disposal rights.133 To sum up Norwegian parties’ positions towards cooperatives, it can be said that right-wing parties accepted the specific role of housing cooperatives in Norway in the first post-war decades (for summary see table 24). However, they changed their positions since the 1970s, when they abolished and amended statutory regulations and cooperative self-regulation rights in favour of individual owners’ disposal rights and, since the late 1990s, they also increased external investors’ influence on cooperatives. The Labour Party, which, in government, had originally shaped the legal framework for cooperative housing, defended the original housing policy framework and the specific role of social-collectively operating cooperatives until the early 1980s. After the late 1980s, however, the Labour Party abandoned their support for cooperatives. This may also have been due to a lack of a left-wing majority for social democratic governments from 1981 until 2005 (for governments and majority compositions, see Appendix A table A21). However, this would not explain their support for the Cooperative Acts in 2003, which increased individual owners’ and external public infringements on cooperative self-regulations. Overall, they seemed to increasingly embrace public housing, which is

160 indicated by their introduction of State Housing Bank loans in 1982 for public housing and further extensions of special housing programs in the following decades.134 In sum, it can be said that the Labour Party changed its former supportive stance towards cooperatives.

•

161 Table 24: Norway

State regulations Political parties1 advocacy Regulations of cooperative tenure Regulations of cooperatives' rights

1946 Price restrictions Superordinate coops' stipulation of sales restrictions allowed All parties 1959 Superordinate coop's stipulation of permanent affiliation allowed DNA 1974, Conversion of affiliated coops into independent coops restricted DNA, SV, 1978 KRF, SP, V 1977 Price and sales restrictions extended to DNA, SV independent coops 1982 -Price limits lifted for affiliated coops; H, KRF, SP abolished for independent coops 162 -Conversion of independent coops into condominiums allowed 1988 Price limits abolished for dwellings older than 7 H, KRF, SP years

1999 Price limits abolished H, KRF, SP

2003 -Conversion of affiliated coops into independent coops freed All parties -Superordinate coops' stipulation of permanent affiliation limited to 10 years -Superordinate coops' stipulation of sales and price restrictions abolished -External investors' right to purchase up to 30% of coop dwellings

1Not included: Communist Party, Progress Party, DNA (Labour), SV (Socialists), H (Conservative Party), KRF (Christian People’s Party), SP (Centre Party)

Summary Subsidy and construction site regulations It has been argued that statutory subsidy regulations supported the growth of cooperatives by giving them relatively more favourable subvention terms (see section 3). From the 1940s until 1982 in Norway and 1992 in Sweden, neither private landlords nor other associational providers gained subventions on the same level as cooperatives. These supportive subsidy conditions were enacted by social democratic governments and reduced by right-wing alliances. However, these subventions for cooperative housing were only controversial in Sweden, where right-wing parties opposed the favourable shaping of subsidy terms and afterwards reduced subsidies against the votes of the social democrats. In Norway, political parties agreed upon the shaping of subsidies for cooperatives to a stronger degree. Only the Conservative Party has shown reluctance to favour cooperatives since the 1950s. During the period of favourable subsidy terms, (affiliated) cooperatives were able to greatly expand their shares within the • housing stock (see table 22). Likewise, construction activity of (affiliated) cooperatives was at high levels under the terms of subvention introduced in the 1940s (see figures 10 and 11). In Sweden, construction activity of affiliated cooperatives exceeded activity of private investors and independent cooperatives until the 1990s. In Norway, affiliated cooperatives could participate in the upswing of construction activity during the 1960s and 70s. Since the 1980s, political parties have partly altered their former stance concerning cooperatives. The favourable subvention terms for cooperatives were reduced in Norway 1982 and abolished in Sweden 1992. The alterations of the once established subsidy conditions were passed by right-wing coalitions, but were not changed by subsequent social democratic governments. After these reductions, cooperatives’ growth within the housing stock slowed or reversed. In Sweden, both private investors and independent cooperatives exceeded affiliated cooperatives’ construction activity since 1992. In Norway, affiliated cooperatives’ construction activity lagged behind that of others overall, including investor categories such as independent cooperatives, public authorities and others.135

163 In addition to subsidy regulations, municipal construction site planning was changed for cooperatives in Norway. A right-wing alliance shifted obligations for municipalities from close collaboration with cooperatives to a more neutral consideration of all investor categories concerning construction site planning in 1982. This could have affected cooperatives’ ability to start housing projects. The Social Democrats opposed these changes, but did not re-introduce formerly favourable access to land for cooperatives.

Operation regulations Besides differences of political parties’ moulding of subsidy conditions, further differences can be seen concerning their interference with internal operations of cooperatives (see also section 4). Parties differed in their willingness either to defend or alter the social-collective operation of cooperatives. Social democratic parties passed legislation incorporating affiliated cooperatives’ price limitations (1940s in both countries), queuing systems (with the aim of dampening price increases, 1960, Norway) and permanent affiliation (1960, Norway) (see tables 23 and 24). When inhabitants started to opt out of social-collective arrangements, social democratic parties defended cooperatives’ regulations via statutory law. They restricted single cooperatives with respect to de-affiliation (1974, 1978, Norway) and legally assured the rights of affiliated cooperatives to regulate prices themselves (1985, Sweden). Furthermore, they opposed infringements on cooperatives’ own regulations (1982, Sweden). Right-wing parties, on the other side, were more ready to oppose and alter the social-collective design of cooperatives. Right-wing parties opposed permanent affiliation of cooperatives (1960, Norway), allowed single cooperatives to dissolve from cooperatives (1982, Norway), forbade affiliate cooperatives from regulating their inhabitants’ selling prices (1992, Sweden) and lifted price limits (1982, 1988, 1999, Norway). However, since the late 1980s in Norway and the beginning of the 1990s in Sweden, social democratic parties accepted past restrictions placed on cooperatives’ social-collective operation. In Norway, the Social Democratic Party even actively voted in favour of abolishing permanent affiliation in 2003. Changes made after the 1980s and 1990s may indicate more profound changes in political parties’ interests in cooperative housing.

164 This is especially the case for the social democratic parties, which supported increased subsidies for rented housing in Norway and Sweden and, in Sweden, introduced a new legal type of associational- cooperative housing. In order to examine social democratic parties’ interests, and changes in their interest in cooperative housing, a closer look at social democratic parties’ and cooperatives’ social and ideological commonalities will be undertaken in the next section.

6. Parties and associations The hypothesis put forward says that parties’ interest in associational welfare provision is based on interests in socially “congruent” associations, that is, associations that have intersecting and coinciding social bases. Associational welfare provision is seen as an opportunity to build and nourish constituencies. It does so by connecting individuals to certain associations that control welfare provision and communicate social norms in ways that are in accordance with parties’ own targets, eventually, also consolidating the political parties’ support bases. • Social Democratic governments supported housing cooperatives until the 1990s in Sweden and Norway. Their support was expressed as influence on state regulations in two areas: they supported cooperatives with favourable subsidy conditions and by allowing them greater independence to self-regulate their provision. This phase will be called Phase I. After the 1990s, in Phase II, the Swedish Social Democratic Party ceased to favour cooperative housing and the Norwegian Labour Party started to support public rented housing as well as cooperative housing. In order to examine the degree of congruency that existed between Social Democratic parties and cooperatives, the social origins, normative orientations and social compositions of party supporters and cooperative tenants are described. Origins and normative orientations The largest cooperative organisations, HSB and Riksbyggen in Sweden, OBOS and NBBL in Norway, are all connected to other labour-based organisations. HSB, for example, was founded by the Swedish tenants’ movement’s Hyresgästernas Riksförbund in 1924. Close connections between Hyresgästernas Riksförbund and HSB persisted over several decades. HSB members were collectively members of the tenants’

165 organisation and they acted together as a pressure group concerning housing policies (Gustafson 1974; Lundevall 1992). Riksbyggen and OBOS136 were likewise founded by labour movement organisations. In both cases, construction worker unions were the initiators (founded in 1941 and 1929, respectively; Annaniassen 1991, 1996a).137 Riksbyggen is further connected to other labour movement organisations, due to direct financial involvement of the national blue-collar union, LO, the national cooperative consumer organisation, KF, and a large labour-based cooperative insurance company, Folksam. There has generally been practical cooperation between the labour movement and cooperatives. HSB, for example, developed and built shops for the central consumers’ organisation Konsum. Further, it contracted insurance at the labour-based insurance company Folksam (Gustafson 1974). In the first decades, housing cooperatives even commissioned worker cooperatives for their construction projects (e.g., HSB and NBBL; Gustafson 1974, Annaniassen 1996a). Furthermore, pressure-group activities and consultation with other labour organisations occurred, for example, via a housing committee of HSB, Riksbyggen and LO or via membership in national cooperative umbrella organisations such as KF, SAMFO or Samvirke.138 Along with inter-organisational labour movement activities, housing cooperatives tried to create a cooperative housing culture among their inhabitants. Cooperatives, for example, invited incoming inhabitants to collectively visit the construction site and discuss the design of planned dwellings, and they encouraged them to take on administrative responsibilities within the local cooperatives. Cooperatives also initiated information evenings and assigned culture committees to organise leisure activities, such as social evenings, sporting events, choirs, amateur theatre, etc. Information about the special character of housing cooperatives was spread through magazines (Annaniassen 1996a, Lundevall 1994). Besides being part of a wider network of labour-based organisations, cooperatives and social democrats shared convictions and aims concerning housing development. On the one side, they agreed that housing should not be privately organised and distributed at market prices, but cooperatively and at cost. On the other side, they supported the notion that “social housing” that is, low-cost housing of

166 a decent standard, should lead to better housing conditions for broader social strata.139 As has been shown above, policies of cooperatives have been to sell their dwellings at cost to incoming tenants, forgoing profits. In addition, they tried to limit selling prices for older dwellings. Furthermore, the main cooperatives in Sweden and Norway are organised as “open” affiliated cooperatives (see section 4). The aim of this kind of organisation was to improve housing conditions for a wider and more inclusive circle of persons.140 HSB’s original aim was “…to act for the lowest possible housing prices and to prevent speculation within housing.” It wanted to be a “permanent working organisation” (Sven Wallander, first HSB director, cited in Lundevall 1992) and to provide “housing with better quality at lower costs.” Riksbyggen wanted to “promote the cooperative form of housing and the right of all persons to good housing.” Guiding principles of the Norwegian umbrella organisation NBBL are “self-help, equity, equality and solidarity.” The biggest Norwegian regional cooperative, OBOS, stated “that housing should be offered to members at cost and that housing should be of ordinary • but good standard.” The purpose has been “to improve city housing conditions and to assure continuous production of housing through a members’ organisation.”141 The social democratic parties were likewise interested in reshaping housing provision away from pure market organisation and improving housing for the less affluent. The Swedish Social Democratic Party regarded housing as “a basic right,” not a commodity. In their outline for post-war politics, Social democratic representatives wrote that “…loans and subventions should support cheaper and better housing provision and non- speculative private interests,” and that “…the speculative character should be removed, which shaped construction activity and building administration.”142 The Norwegian Labour Party formulated the aim that “all working Norwegians, even those with the lowest incomes, should have an adequate dwelling at a price they can pay.” Furthermore, the Norwegian Labour Party “wills not that market forces alone are the basis for the housing area,” and that “…the State Housing Bank shall not finance uncontrolled and speculative construction activity.”143 In sum, social democratic parties could regard the large cooperatives as organisationally and normatively sympathetic

167 organisations. Both groups had a common social origin within the labour movement and a common commitment to alter housing conditions and the housing market. Additional efforts by housing cooperatives to create a notion of “solidarity and community” among inhabitants can be assumed to be compatible with the interests of the Social Democratic parties. Today, however, cooperation with other labour movement organisations has weakened. The housing committee of HSB, Riksbyggen and LO no longer operates, for example, and HSB abolished collective membership in the national tenant organisation in 1985.144 Moreover, the aims of cooperatives became rather unspecific. HSB, for example, wants “to create good housing in cooperation” and its core principles include “commitment, security, sustainability, care and cooperation” (HSB). Instead, cooperatives highlight their “good reputation” as housing developers, the security they offer as large organisations or their “attractiveness,” due to “good services” and “good quality” of housing: “One has to be competitive at any time in a market economy. It is important that cooperatives have a supply, which members want at any time” (interview with NBBL director Ralph Norberg, in Bo; HSB, Riksbyggen, OBOS, NBBL).145 OBOS describes its situation in the following way: “OBOS’ role has changed since the Norwegian housing market was deregulated in the 1980s. From being a body that carried out public housing policy, the organisation today is now exposed to competition and works under the same basic conditions as the rest of Norwegian business and industry” (OBOS).

Constituencies Cooperative owners’ socio-economic characteristics An early study of cooperative owners of HSB Stockholm shows that 75% were workers and salaried employees in 1939. Workers belonged predominantly to the worker “aristocracy,” that is, they were skilled workers such as craftsmen and construction or metal workers. White- collar employees were predominantly salaried public sector employees: railway, public transport and postal employees, clerks and caretakers (Gustafson 1974). HSB thus accommodated not the poorest strata in Stockholm but rather middle classes with some form of secure and average-level income.

168 Also, in later decades, cooperative tenants’ socio-economic composition mirrored the average socio-economic composition of the population. In 1960, predominantly middle-income salaried employees owned cooperative dwellings. Besides employees at middle-income levels, higher-salaried and lower-salaried employees and workers were the main groups living in cooperative dwellings (see table 25). Compared with other tenure households’ social position, cooperative occupants can be said to have a middle position between home occupants and renting tenants.

Table 25: Sweden: Households by tenure and socio-economic status, percent, 1960

All Cooperative Owner- Renting households tenants occupiers tenants, others

Farmers, fishers, foresters, etc. 11 0 23 5 Self-employed 7 6 10 6 Higher salaried 2 2 1 2 Middle salaried 22 34 11 26 Lower salaried 2 2 1 4 Workers 32 36 28 33 Others 24 20 26 24 • All 100 100 100 100 n 257904 24219 93457 140228 Source: SOS Population and Housing Census 1960

Income levels for tenants in cooperative dwellings have likewise been at an average level for the population (Gustafson 1974, Lundevall 1996). In 1990, for example, the average household income for cooperative owners was 199 100skr, while the average for all households was 220 200 skr (owner-occupiers had 288 800 and renting households 167 800skr; SOS Population and Housing Census 1990). In Norway,146 the composition of cooperative tenants is roughly similar to the Swedish composition. Salaried employees as well as workers are the typical socio-economic groups living in cooperative dwellings (see table 26 and Appendix A table A34 for 1995). The household income of cooperative tenants lies usually around the average income (see Appendix A table A35). However, the income level of cooperative inhabitants seems to have decreased between 1973 and 1994: it was slightly above the average income in the earlier decades and slightly below in recent decades.147

169 Table 26: Norway: Households by tenure and socio-economic status, percent, 1981

All Cooperative Owners Renting households tenants tenants, others Farmers, fishers, foresters, etc. 5 0 8 2 Self-employed 10 7 12 6 Higher and middle salaried 18 20 19 15 Middle and lower salaried 12 15 11 12 Workers 21 19 24 16 Others (other economically 34 39 27 48 active, economically inactive) total 100 100 100 100 n 2201 358 1300 543 Source: NOS Survey of Housing Conditions 1981

In sum, Swedish cooperative inhabitants mirror the average of distribution of socio-economic and income groups within the whole population. Congruency with social democratic parties’ social background is thus not self-evident due to the socio-economic composition of cooperative households. Norwegian cooperative inhabitants first consisted mainly of middle-class groups at slightly above-average income levels. However, in the 1990s, they consisted more of lower-income groups.

Cooperative owners’ political support Sweden Election studies show, however, that people living in cooperatives were strong supporters of the Social Democratic Party in the 1970s (see table 27; in earlier election studies, dwelling ownership was not specified). A 53% majority of the interviewed cooperative owners voted for the Social Democratic Party, while only a plurality of all interviewed did so. In addition, cooperative owners’ voting for the Social Democratic Party exceeded Social Democratic voting by all other tenant types. It was stronger even than that of renting tenants, exceeding average voting by 9 percentage points compared with 7 percentage points by renting tenants (see Appendix A table A36 for percentage differences). Likewise, the relative difference between cooperative occupants’ social democratic voting on one side (53%) and right-wing voting on the other side (39%) was rather large, with a difference of 14 percentage points, while of all persons questioned voted more often for right-wing parties than for the social democrats. Furthermore, cooperative owners exceeded average voting for the Social Democratic Party, with 9 percentage points, while cooperative

170 owners’ voting for right-wing parties remained 7 percentage points below average voting. The strong support of cooperative owners, nevertheless, disappeared in later decades. In 2002, cooperative owners no longer voted more for the Social Democratic Party than for other parties (see table 28 and Appendix A table A38 for percentage differences). The Liberal and Conservative Parties, especially, gained votes from cooperative owners. In addition, differences between cooperative owners’ voting and other tenants’ voting had vanished. Over the course of several national elections there has been a gradual decline in cooperative owners’ social democratic voting, both in total and in comparison with average voting for the Social Democratic Party since the 1970s (see Appendix A figure A5).

Table 27: Sweden: Voting and tenure, national elections, %, 1976 SV* SAP C FP M N

Multi- (Co-op-)1 3 53 14 14 11 254 dwelling owned Rented 6 51 16 10 10 869 Detached Owned 2 37 28 12 16 1159 • Rented 3 47 26 6 11 144 Other2 5 28 36 11 13 100 Total 4 44 22 11 13 2526 1Single dwellings in Sweden cannot be owned individually 2 old-age housing, facility housing, etc. *SV (Left Party), SAP (Socialdemocratic Party), C (Centre Party), FP (Liberal Party), M (Conservative Party) Source: SSD Swedish Election Studies

Table 28: Sweden: Voting and tenure, national elections, %, 2002 SV* SAP C FP M N

Multi- (Coop-)1 9 34 3 20 12 317 dwelling owned Rented 11 35 2 12 8 597 Detached Owned 6 35 7 13 12 1248 Rented 8 30 7 7 12 84 Other2 14 28 3 11 3 36 Total 8 34 2 14 11 2282 1Single dwellings in Sweden cannot be owned individually 2 old-age housing, facility housing, etc. *SV (Left Party), SAP (Socialdemocratic Party), C (Centre Party), FP (Liberal Party), M (Conservative Party) Source: SSD Swedish Election studies

171 Norway

Strong Labour support among cooperative owners is indicated in Norwegian national election studies (unfortunately, cooperative tenure is not specified in the national elections before 1969 and after 1981). Cooperative occupants were the most supportive group for the Labour Party in the national elections of 1969 and 1981 (see tables 29 and 30). In 1969, a majority of cooperative owners voted for the Labour Party. In 1981, the support was diminished, but still many cooperative owners voted for the Labour Party. The difference between cooperative owners’ voting support for the Labour Party and for other parties was very high in 1969, but decreased in 1981: it fell from +33 to +10 towards the conservatives and from +38 to +18 towards other parties. However, these reductions do not apply to cooperative owners alone. In other words, overall voting support for the Labour Party diminished and renting tenants reduced their support even more strongly. When comparing with other types of tenure, it can be seen that cooperative owners voted for the Social Democratic Party more often than other occupant groups in both elections. The difference in average occupants’ voting was only slightly reduced from +8 to +5 percentage differences from 1969 to 1981.148 In both elections, social democratic support among cooperative owners was high.149

Table 29: Norway: Voting and tenure, national elections, %, 1969 Percentage-differences from average DNA* %-% H %-% Others %-% N

Cooperative 57 8 24 -7 19 -5 105 Owners

Renting tenants 54 5 16 -1 30 -4 433

Owner-occupiers 45 -4 16 -1 39 5 757

Total 49 17 34 1295

*DNA (Labour), H (Conservatives) Source: Bay 1985

172 Table 30: Norway: Voting and tenure, national elections, %, 1981 Percentage-differences from average DNA* %-% H %-% Others %-% N

Cooperative 43 5 33 -3 25 -2 216 Owners

Renting 32 -6 35 -1 33 6 232 tenants

Owner-occupiers 38 0 37 1 25 -2 892

Total 38 36 27 1340

*DNA (Labour), H (Conservatives) Source: Bay 1985

Summary Social democratic parties and cooperatives shared a common social origin in the wider labour movement, from which they received organisational and financial support and to which they held elite connections (tenant movements, unions, insurance companies, etc.). Parties and associations shared similar socio-political aims for • developing housing in the early decades. Both attempted to shape the functions of housing away from a speculative free market. In engaging in cooperative housing, one could say that they attempted to establish a new type of ownership, connecting the security and housing standards of individual ownership with social-collective access and disposal limitations. In addition, cooperatives and social democratic parties had rather congruent constituencies. Even if cooperatives addressed middle and not lower socio-economic classes, their members and occupants belonged to the constituency of the social democratic parties until the 1980s. In recent decades, however, cooperatives and social democrats seem to have developed away from this original set-up in Norway and Sweden. Both reduced their ambitions to diminish market influences on housing provision. Both also loosened their connections with other labour-movement organisations, although to varying degrees. While social democratic parties weakened their connections, the largest cooperatives dissolved theirs with other labour organisations.

173 The development of constituencies shows a somewhat diverging pattern in Norway and Sweden: In Norway, the share of lower classes increased slightly, while in Sweden the socio-economic composition of cooperative tenants seems not to have changed. The electoral support of cooperative tenants for the Social Democratic Party is no longer above the average of that given by other types of tenants in Sweden (such data are not available for Norway). The subsequent strategic choices of parties and associations diverged slightly between Norway and Sweden. In Norway, the Social Democrats still support cooperative housing, but have complemented their housing strategies by subsidising public rented housing for weak groups on the housing market (e.g., students, pensioners, immigrants). In Sweden, the social democrats completely stopped favouring cooperative housing in its traditional form, but opted instead for policies favouring cooperative tenantship and public housing for multi-dwelling houses. Cooperatives, on the other side, opted to survive as providers of a condominium-like, market disposed tenure in competition with private providers and independent cooperatives in Sweden and Norway. In sum, the original labour movement constellation of parties and cooperatives weakened in Norway and dissolved in Sweden. From being part of a wider constellation of labour movement organisations, parties and cooperatives today have to survive more independently from each other in electoral and housing markets. This development has been more profound in Sweden. The hypothesis that partisan interest in associational welfare provision is based on congruent constituencies is partly supported. Social democratic parties gave strong support to cooperatives as long as they shared congruent constituencies with the main affiliated cooperatives, but they either weakened their support (in Norway) or fully abandoned their preference for them (in Sweden) when the congruency with cooperatives weakened. Right-wing parties, on the other side, were less supportive, while sharing fewer intersecting and overlapping constituencies with cooperatives. Furthermore, right-wing parties’ support for individual ownership rights of cooperative tenants seems to have coincided with an increased share of votes from cooperative tenants for those parties. In addition, it coincided with cooperatives’ increased free-market orientation since the 1990s in Sweden. These indicate an increase in

174 congruency between right-wing parties and cooperatives, which would be in accordance with the presumed hypothesis.

7. Housing: Summary of main findings The hypothesis here is that statutory support for welfare associations is moulded by political parties in accordance to their particularistic interests in welfare associations. Political parties are supposed to attempt to influence both statutory subvention of and regulation of welfare associations and, thus, to attempt to influence the level of associations' welfare production and the type of welfare, associations provide for. Furthermore, political parties’ diverging interests in associational welfare provision are supposed to be based on whether the parties share more or less “congruent constituencies” with associations. The examples discussed are Norwegian and Swedish parties’ politics concerning housing associations.

• I. Partisan policies Parties diverged in their policies concerning two sets of statutory legislation: subsidy distribution and operational regulations. First, social democratic governments established subsidy systems that were more generous for cooperatives than for private developers, other types of associations and, in Norway, public authorities. These favourable subsidy systems were supported by right-wing parties in Norway, while, in Sweden, right-wing governments first opposed favourable subsidy regulations and later reduced subsidy allocations for cooperatives. Second, social democratic parties proposed and shaped operational regulations in favour of cooperatives’ statutes. They either incorporated former cooperative statutes into statutory law (e.g., price regulations) or assured via legislation cooperatives’ rights to regulate their operations independently (e.g., cooperatives’ regulations concerning access to and disposal of their dwellings). In contrast, right-wing parties opposed or reduced cooperatives’ self-regulation and incorporation of laws. Instead, they enhanced occupants’ rights to dispose of cooperative dwellings individually (e.g., to sell dwellings at market prices, to transform them into condominiums or to choose their buyer).

175 Due to their influence on government regulations concerning statutory housing intervention, social democratic parties could shape conditions selectively, in ways that favoured cooperatives over other developers of multi-dwelling housing. As a result, social democratic politics strengthened the growth of cooperatives in comparison with other developers and strengthened their specific social-collective design (see summary of section 5). Since the late 1980s, social democratic parties have reduced their support for cooperatives. On the one side, they no longer shape subsidy conditions to favour cooperatives. Instead, they increased subsidies for public housing. On the other side, they no longer attempt to shape legislation to enhance the social-collective character of cooperative housing, but have accepted or supported legislation that reduced the distinctive operation of cooperatives.

II. Parties’ and associations’ constituencies The variation of political parties’ support for conditions that favoured or disfavoured cooperatives is supposed to be based on interests in socially more or less congruent associations. The examples examined are the large housing cooperatives, which became the predominant form of cooperatives, and social democratic parties, which supported cooperatives via subsidies and operative regulations in Sweden and Norway. Evidence suggests that parties and cooperatives shared interests in social collective housing provision as well as overlapping constituencies and connections with the wider labour movement. First, owners of cooperative dwellings were above-average voters for social democratic parties in Norway and Sweden. In other words, the same individuals belonged to the social bases of both cooperatives and social democratic parties. Second, both parties and cooperatives had elite connections with other labour organisations. Furthermore, both received organisational or financial support through other labour organisations (tenant movements, unions, labour-based insurance companies, etc.). Third, housing cooperatives shared interests with social democratic parties in housing. Both wanted to reduce the influence of the free market and considered decent housing to be a social right. Both engaged in cooperative housing as an alternative that could provide a form of social-collective owner-occupation.

176 These similarities in social background are here interpreted as underlying social democratic parties’ modification of statutory regulations in favour of cooperatives. By contrast, right-wing parties’ less supportive politics towards cooperatives is regarded as being due to an absence of social links between right-wing parties and the large cooperatives. In the 1990s, parties’ and cooperatives’ constituencies and housing aims were no longer as congruent as they had been in the early decades after the Second World War. The social composition of cooperative tenants seems not to have changed, but electoral support of cooperative tenants for the Swedish Social Democratic Party is no longer above the average of other tenants. Furthermore, social democratic parties and cooperatives diverge concerning their aims and connections with other labour movement organisations. While both reduced their ambitions to diminish market influences on housing provision, the strategic choices seem to differ: Social democratic parties are still to some degree interested in subsidising social housing for weak groups on the housing market (e.g., those in small dwellings, students, pensioners), • while cooperatives have opted to survive as providers of a market- disposed tenure in a competitive market. Decreased social congruency between social democratic parties on one side and cooperatives on the other are here interpreted as one reason why social democratic parties stopped favouring and defending cooperative housing by means of statutory regulations. In sum, the interpretation that party interest in associational provision is not based on principal positions on welfare organisation, but on interests in socially congruent associations, is supported by two variations. First, social democratic parties supported cooperative housing, while right-wing parties maintained a more distanced position. Second, the relevance of congruent support bases is indicated in the reduction of social democratic support for cooperatives in Sweden after the 1990s. Diminished electoral support of cooperative tenants for the Social Democratic Party was followed by reduced interest from party. A further variation of political parties’ interest in associational welfare provision is the different stance of parties concerning associational welfare in housing, on one side, and day-care, on the other side. While social democratic parties were the main political

177 supporters of cooperatives as housing providers, they were more distanced towards associational provision of day-care. Right-wing parties, on the other side, which pursued delimitation of cooperatives’ independence and, in Sweden, reductions to housing subsidies, were more supportive of associational provision of day-care.

178 IV. Discussion and Conclusion

1. Introduction The aim of this thesis was to examine whether political party preferences for certain welfare associations might help to explain variations of statutory support for associational welfare provision. For this purpose a comparison was made between political parties’ attempts to influence statutory regulation of and subvention of Norwegian and Swedish associations active in the areas of day-care and housing. The main questions were if parties differed in their attempts to influence statutory subvention and regulations of associational welfare provision in order to support or to restrict associational welfare activities and if partisan differences could be understood as based on interests in associations with similar policy goals and similar supporter segments. It was hypothesised that parties • will pursue diverging policies in order to support association-friendly policies in regard of associations with congruent constituencies. The hypothesis, which emphasises to a greater degree the role of political parties in shaping statutory policies and broadens the theoretical basis of partisan interests in associational welfare, was intended as a refinement of current attempts to explain differences in associational welfare provision (see introduction). In order to discuss the empirical findings out of various hypothetical perspectives, I will, in the following, first, focus on divergences and congruencies between political parties concerning their efforts to shape subventions and regulations for associations. In a second step, I will turn to a discussion of the here proposed motivation of political parties to be more supportive for associations with which they have congruent constituencies.

2. Partisan policies concerning welfare associations Many scholars assume that right-wing parties tend to support associational welfare provision per se, while left-wing parties are

179 supposed to prefer public welfare provision per se (Salamon and Anheier 1998) Instead, the hypothesis here suggested claims that political parties diverge concerning their policies towards associations, according to their varying interests in certain associations. The findings in the foregoing chapters show that social democratic parties as well as right-wing parties supported associational welfare provision via statutory subvention and regulation. Furthermore, social democratic parties as well as right-wing parties sought to influence associational welfare provision. Norwegian and Swedish social democrats both issued association-friendly policies concerning the subvention and regulation of housing cooperatives, while they pursued unfavourable regulations and subventions for associations active in day-care. Right-wing parties, on the other side, pursued association- friendly policies for day-care associations, while restricting regulations and subventions that favoured housing cooperatives. Thus, former approaches, which suggested a general ideological stance towards associational welfare provision – a distanced stance by social democratic parties and an association-friendly one by right-wing parties – are not regarded as being in accordance with the observations made. Instead, political parties pursued association-friendly policies for some associations, but not for all forms of associations. Other scholars claimed that associational welfare provision developed in a consensual way in as a “partnership” between states and associations (Kuhnle and Selle 1992). Evidence supporting such an understanding could be seen in cases when political parties consensually agreed upon subvention and regulation of associations. This was, for example, the case when Norwegian right-wing parties supported the introduction of generous statutory subventions for housing associations and when Norwegian social democratic parties in principle accepted subventions for day-care associations, regarding associations as a “supplement” to public day-care. If one considers the importance of statutory subvention for associational welfare activity, this can be regarded as partial evidence for a partisan consensus in Norway. On the other side, Norwegian political parties often pursued their associational interests in a conflicting way (Weisbrod 1990, James 1987). The Norwegian Labour Party delimited subventions for some day-care associations and attempted to restrict day-care associations’ relative freedom to select personnel, determine curricula and select clients. These measures were met by resistance from Norwegian right-

180 wing parties (and affected day-care associations). Further examples of partisan conflicts are Norwegian left-wing and right-wing parties’ decreeing of regulations concerning cooperatives against the opposition from either right-wing or left-wing parties (and against the wishes of housing cooperatives). In Sweden, parties pursued conflicting policies concerning both relative subvention levels and regulations. Swedish right-wing parties, for example, voted against the introduction of generous subsidies for cooperatives and abolished cooperatives’ preferential position concerning subsidies. Furthermore, they restricted housing cooperatives’ control over dwelling-sales. Swedish left-wing parties, on the other side, delimited day-care associations’ eligibility for statutory subventions and introduced maximum levels for parent fees. In sum, the understanding of a consensual development concerning associational provision is regarded as underestimating underlying conflicts about statutory interventions into associations’ activities. The politics of Swedish parties are regarded as more in line with the assumption of partisan conflicts concerning the involvement of associations than is the case in Norway, where, parties across the left- • right divide (partially) agreed on subsidising associations. A further approach assumes that ethnic or cultural heterogeneity leads to higher levels of associational welfare provision (Weisbrod 1990, James 1987). Underlying this presumption is the assumption that competitive ethnic and cultural minorities will be eager to provide welfare in order to attract adherents. Among the observed cases in this thesis, the higher activity level of religious-conservative day-care associations until the 1990s in Norway can certainly be accounted for by an understanding that cultural heterogeneity influences associational welfare activity. Religious-conservative day-care providers can be said to have been organised by a cultural minority in Norway, and they were eager to provide certain forms of day-care (i.e. with religious curricula, for older children and with short opening hours).150 Likewise, the greater interest of associational providers of day-care in Sweden since the 1980s could be understood as being based on an increase in culturally more diverse day-care interests, not accounted for by entirely publicly organised provision. A wide range of small associations became active, the most prevalent organisation types being parent cooperatives and individuals, but also including employers, Waldorf and Montessori associations and religious

181 organisations. The predominance of various small types of associations seems to reflect quite heterogeneous “micro-interests,” ranging from parents’ self-organisation to pedagogical and religious interests. However, some doubt remains about an interpretation that the increase of associational day-care was fostered by an increase in culturally and ethnically heterogeneous day-care interests. First, public and associational day-care provision became rather similar in both countries in terms of ages of children attending and opening- hours. In Norway, this implied that operations of associational day- care provision became less heterogeneous. Second, only some Swedish parents expressed a specific interest in associational provision when asked about their motivation for sending their children to non-public day-care centres. A positive interest by parents was stated only by parents of children in parent cooperatives, while other parents stated mere practical reasons for choosing non-public day- care, such as, proximity to workplaces or homes. A further difficulty with applying the heterogeneity approach to welfare activities, as observed here, is that it would be difficult to understand the activity of labour-movement-related housing cooperatives in both countries as predominantly motivated by cultural interests. Instead, their main motivation could be seen as socio- economically motivated. Thus, the activity of housing cooperatives would not be accounted for by an approach that focuses solely on cultural and ethnic heterogeneity. In order to include labour-related or other socio-economic interests as motivations for associational activity, the heterogeneity approach would have to be widened to embrace the assumption that diverging particularistic interests per se can be regarded as motivations for associational activities. The main objection to the heterogeneity approach is, however, the assumption that minorities with particularistic interests would be able to provide welfare without statutory subvention. Statutory subvention has been an important financial resource for welfare associations; conversely, statutory regulations have also constricted associations’ welfare operations. This implies that political parties had the opportunity to influence the activity level and operations of welfare associations. The question is, did parties attempt to influence associational activity by altering statutory conditions for associational provision, or has social particularism/heterogeneity been a straightjacket for partisan policies? Here, it is argued that partisan

182 interests have pursued conflicting policies, especially where the welfare interests of associations were regarded as “too particularistic.” Parties have hereby been influential in shaping the activity level and operations of associations, even in conflict with other parties (and associations). Concerning, for example, the growth of Norwegian religious- conservative day-care associations, it is argued that generous statutory subvention for short-time day-care catering to older children provided them with financial advantages in comparison with other associations. In addition, ample statutory regulation of day-care operations gave these associations freedom to operate according to their cultural inclinations (e.g., concerning personnel selection, curricula, client selection) and, thus, was an incentive for them to become active. The relative decline of religious-conservative day-care associations in comparison with other day-care providers since the 1980s and 90s is here argued to have been brought about in part by a lowering of subsidies for short-time day-care and care for older children. These preconditions were, in turn, subject to partisan conflict. On one side, right-wing parties, especially the Christian People’s • Party, defended and maintained favourable conditions for religious- conservative day-care providers. On the other side, the Labour Party attempted to cut subsidies and to delimit operational freedom for, in particular, religious- conservative day-care associations and successfully decreased subventions for short-time day-care and day-care catering to older children. In Sweden, the growth of day-care associations was enabled by statutory subvention, but arguably hampered by delimited access to short-time day-care subsidies, preclusion from extra subsidies available for municipal providers and restrictions of eligible categories of associations. These access restrictions may have conflicted with associations’ predominant activity in short-time day-care and may have reduced their chances for establishing themselves alongside municipal providers. Access restrictions for associations were decreed by social democratic governments, whereas gradually enhanced access to public subvention, on the other side, was pressured for and then exhaustively institutionalised by right-wing parties. In addition, social democratic parties introduced maximum levels for parent fees against the opposition of right-wing parties, which may conflict with the

183 interests of any associations wishing to offer exceptionally high- quality day-care for more affluent clients. Considering also whether housing cooperatives are affected by particularistic interests, it is here argued that statutory subvention and regulation has been favourable for the relative growth of cooperatives as well as for their specific social-collective operation in Sweden and Norway. These favourable conditions, however, became subjects of partisan conflicts, especially concerning the social-collective conception of cooperative housing until the 1980s/1990s. Social democratic parties, on one side, introduced statutory assurance of cooperatives’ social-collective operation, which was met by the opposition of right-wing parties in both countries. Right-wing parties, on the other side, later transformed cooperative housing from social-collective agencies towards more market-like operators against the opposition of left-wing parties (and cooperatives) in both countries. Additionally, in Sweden, right-wing parties opposed favourable subsidy regulations, introduced by a social democratic government in the 1940s. Right-wing parties later abolished the older subsidy system and equalised the eligibility of cooperatives with that of private investors in multi-dwelling housing. In Norway, however, subvention of housing cooperatives was not contested between parties and remains today more favourable for cooperatives than for private investors in multi-dwelling housing. Thus, partisan conflicts emerged precisely around associations’ “particularistic” welfare provision. An understanding that particularism/heterogeneity has delimited the efforts of parties to pursue conflicting policies is, therefore, not regarded as entirely convincing. When summarising the arguments for and against various hypothetical approaches for understanding associational welfare provision, it could be assumed that none of the presented approaches, including the hypothesis I have proposed, accounts for all observed variations of level and type of associational welfare provision. Partly, associational provision was favoured by political party consensus about their involvement, an indication that parties regarded associations per se as “partners” for providing welfare. Norwegian parties could be argued to have pursued partially such a consensual approach, when agreeing to subsidise most of the active associations. However, they pursued conflicting policies concerning the regulation

184 of associations’ welfare operation. How to explain the more consensual support for associations in Norway remains an open question. The heterogeneity approach may provide an interesting perspective (see below). Partly, associational provision has been influenced by conflicting interests of political parties, in accordance with the hypothesis here supposed. Divergent partisan policies are here argued to have influenced, first, the provision level of (some) day-care associations in Sweden and Norway and, additionally, of housing cooperatives in Sweden. Second, divergent partisan policies are regarded to have been crucial for shaping associational independence to control their operations and for shaping statutory regulations in accordance with associations’ preferences. In the case of housing cooperatives, partisan influence on statutory regulations sustained the character of cooperatives as social-collective housing providers until the 1980s/90s and, afterwards, influenced their shift to market-oriented providers. In the case of day-care associations, partisan influence on regulations assisted religious-conservative associations to provide day-care until the 1990s and, afterwards, contributed to their decline. Hereby, left- • wing as well as right-wing parties supported or restricted associational welfare. Partly, the heterogeneity approach focuses on the important aspect of differences in associations’ activities due to particularistic welfare interests. While housing cooperatives were active in both countries, levels of activity of day-care associations differed between the two countries: day-care associations were nearly absent in Sweden until the late 1970s, although associations were eligible for subventions. By contrast, day-care associations in Norway were very actively using the opportunity to receive subsidies. In a further step of argument, one could combine the particularism/heterogeneity approach with the partnership approach. Heterogeneous welfare interests, indicated by associations’ activity levels, may have influenced partisan consensus over allowing associations to be responsible for welfare provision. Partisan consensus could be understood as the acceptance that compromises are needed to gain broad popular support for welfare policies, which can only be satisfied by involvement of particularistic associations. However, the impression is rather that parties pursued conflicting policies, especially when associations were considered “too” particularistic (see above). Furthermore, cross-partisan acceptance of

185 associational day-care provision continues today in both countries, although this does not necessarily indicate greater partisan consensus on allowing enough elbow room for associations to operate according to particularistic interests. Rather, partisan consensus seems simultaneously to include delimitation of associations’ influence on welfare operations, by more strongly regulating associational providers in favour of public control and/or client rights, thus delimiting the specificity of associational provision. In Sweden, for example, cross-partisan acceptance of equal terms for subsidising associational or municipal day-care since the middle of the 1990s includes partisan agreement upon tight regulation of associations’ day-care, making it part of municipal day-care plans, which regulate client access, age categories attending, opening hours, etc. In Norway, in 2005, parties agreed upon day-care associations’ subjection to municipal coordination of client uptakes, opening hours and ages of children attending. In Norway since the 1980s, by partisan consensus, housing cooperatives have received lowered subsidies in comparison with public providers, thus enhancing comparatively the position of public provision. Also in Norway, public control has been strengthened concerning cooperatives through the enhanced right of municipalities to purchase cooperative dwellings in order to own and, thus, dispose of them (public investors are allowed to purchase up to 30% – up from 10% – of the dwellings within a cooperative). In both countries, all parties accept regulations of housing cooperatives, which imply cooperatives’ loss of the right to delimit the disposition of dwellings by inhabitants and increased freedom for individual owners. In other words, individual owners’ rights have been strengthened against the social-collective rights of cooperatives. In summarising these tendencies of public interference with associational welfare provision, one could speak of a “squeezing” of associations between individual client rights and public control. Thus, partisan consensus over associational involvement does not necessarily indicate that this was combined with partisan hesitation to break with particularistic welfare interests. The influence of particularistic welfare interests on statutory welfare developments is, therefore, regarded as delimited. Further interpretations could be thought of for understanding partisan consensus for subsidising associations.

186 In the case of cross-partisan support for associations, the interest of parties could be understood as a “raison d’etat” acceptance of associations’ ability to provide welfare more efficiently, for example, by using less tax revenues than public providers, or socially effectively, for example, by reaching a broader spectrum of welfare clients than public providers. Day-care associations, eventually, were able to provide day-care more efficiently, for example, by attracting volunteers. Hereby, the activity level of day-care associations may have led to stronger cross- party awareness of associational advantages in Norway than in Sweden. In Norway, parties also accepted that associations demanded higher parent fees than municipal providers until 2005, in order to meet the accelerating demand for day-care under budget constraints (see, e.g., St.meld. nr. 83 1981-82, Brundtland). Norwegian parties risked the possibility that not all social strata would be able to pay for and, thus, to use publicly subsidised day-care provided by associations. Swedish parties, on the other side, may have been more interested than Norwegian parties in the effective provision of day-care until the • 1980ies, whereby Swedish parties regarded public provision as better able to include all social strata. Growing budget deficits since the 1980s in Sweden may have increased cost awareness and efficiency interests, spreading cross-party acceptance of subvention for day-care associations as a necessary supplement to meet increased demands for day-care since the 1990s. In both countries today, partisan consensus reigns over equal terms in subsidising associational or municipal providers. However, in both countries this is combined with detailed municipal control, regulations that could be interpreted as statutory efforts to ensure effective and subsidy-efficient associational welfare production according to statutory interests. A consensus across parties for subsidising housing cooperatives only existed in Norway, while Swedish parties instead consensually voted for generous subsidies for public rented housing in the 1940s. It is difficult to assess whether the Norwegian consensus was based on deliberations to provide “social” housing more efficiently via cooperatives due to the usual standard differences between rented dwellings and cooperative dwellings. Ceteris paribus, however, cooperative housing demands a lower share of subsidised capital from states than rented housing, due to the requirement that inhabitants make a capital investment. This could have been an incentive for parties to opt for cooperative housing in

187 order to save tax revenues in Norway. The social effectiveness of cooperative housing, on the other side, could be regarded as lower than that of rented housing, precisely due to the capital investment requirement. In Norway, the difficulty for certain groups (e.g., young people, people with lower incomes) to make a capital investment led, in the 1980s, to a cross-partisan agreement to enhance public ability to buy cooperative dwellings in order to rent them out, and to provide more subvention for public housing. Irrespective of whether one interprets partisan consensus concerning associations as based on particularistic/heterogeneous welfare interests, efficiency advantages or more or less predominant interests in effective (public) provision, an underlying trend since the 1980s seems to be that parties to an increasing degree agreed upon public regulation, which led to a strengthening of either detailed public control or individual client rights, while associations’ ability for self-governance was weakened since the 1990s. Partisan conflicts, on the other side, seem to have evolved from conflicts about particularistic/heterogeneous associational efforts, until the 1980s, into conflicts about the effectiveness of welfare provision. While partisan conflicts around religious-conservative day- care associations and social-collective housing cooperatives were at least partly about the relative influence of those associations to regulate their own operation, partisan conflicts after the 1990s were about maximum limits for parent fees for associational day-care. However, as there were only a few partisan conflicts after the 1990s, this “conclusion” is probably premature. Which motivations underlie partisan conflicts concerning associational welfare provision? Motivations for diverging and changing partisan policies towards associations are discussed in the next section.

3. Partisan interests in welfare associations The proposed hypothesis suggests that parties will pursue more or less association-friendly policies according to their interests in supporting specific welfare associations. Parties are assumed to aim at supporting associations with “congruent constituencies,” meaning associations with which they share overlapping and intersecting social bases. Partisan conflicts about associational welfare involvement could in turn be interpreted as being based on parties’ relative congruency with

188 those associations that are active in a specific welfare field. Related empirical questions would thus relate to whether parties shared a congruent constituency with those associations active in a welfare area. Other interpretations of partisan motivations to subsidise associational welfare could be thought of. If one were to simplistically characterise parties along a continuum from left to right, where right- wing parties would be assumed to be liberal, pro-market parties and left-wing parties to be socialist “decommodification” parties, then their interest in associational welfare could be assumed to be based on associations’ ability to produce welfare more or less efficiently or effectively than other welfare operators. Right-wing parties could be assumed to be interested in associational welfare involvement out of the hope for tax savings due to associations’ efficiency advantages and/or due to their closer-to- market operation, involving higher individual payment responsibilities instead of tax revenues. In other words, they would, in tendency, rather risk the social effectiveness of welfare provision by associations than risk higher tax payments due to “inefficient” welfare production. • Left-wing parties, on the other side, could be simplistically characterised as more interested in effective associational welfare provision, ensuring the inclusion of less-well-off social strata, which may not be able to pay higher fees in order to receive (higher-quality) welfare. Left-wing parties would, in other words, rather risk higher tax subsidies for welfare production than risk the effective inclusion of less well-off social strata. Thus, partisan conflicts over subsidising and regulating associational provision would occur along a left-right pattern, where left-wing parties would pursue associational provision according to social-inclusiveness deliberations, while right-wing parties would pursue associational subvention according to tax-saving deliberations. Partisan conflicts over day-care associations could be interpreted as conflicts over the relative balance of efficiency and effectiveness interests concerning associational provision. In order to present the findings, I will first briefly present partisan positions on associational welfare provision and then present the main active associations. Thereafter, I will discuss partisan policies towards associations, both in relation to associations’ social bases and in relation to effectiveness and efficiency gains attained from

189 associational welfare production. The question is how to interpret the diverging and changing partisan policies towards associations. As has been described in this thesis, social democratic parties pursued association-friendly policies concerning housing cooperatives, while tending to restrict day-care associations’ subvention and operational independence (concerning client selection, personnel selection, curricula setting and fee setting) until the 1990s. Since then, social democratic parties, especially Swedish social democrats, have opted for stronger support for public housing and have accepted associational day-care. Right-wing parties, on the other side, tended to restrict subventions for housing cooperatives (in Sweden) and pursued a strengthening of individual rights of cooperative inhabitants. Instead, they pursued association-friendly policies concerning day-care associations. In Norway, right-wing parties defended generous subsidies for short-time day-care and day-care for older children. Furthermore, they legally assured associations’ rights to select personnel, curricula and clients according to associations’ preferences and to decide on parent fees. In Sweden, they decreed eligibility for all forms of associations for statutory subsidies and, in addition, made associations eligible for municipal subsidies.

Housing cooperatives The main active housing cooperatives in Norway and Sweden, HSB, OBOS, Riksbyggen, etc., were founded by either tenant organisations or construction-worker unions and could thus be said to have originated in the wider labour movement. Their original aims were to provide housing “at cost” as an alternative to the “speculative” housing market dominated by private landlords. In order to affirm a long-term at-cost orientation, they restricted inhabitants’ ability to sell their apartments freely on the housing market by imposing delimitations either on prices or on inhabitants’ choices among interested buyers. In Norway, they also pressured the state to additionally apply price restrictions. However, such sale delimitations were gradually abandoned after the 1970s in both countries due to pressure from inhabitants and statutory regulations. Today, sales restrictions are almost nonexistent in both countries, making ownership of cooperative dwellings very similar to private, individual home-ownership. The superordinate cooperative organisations, HSB,

190 OBOS, etc., currently act more as market-oriented service providers than socially inspired, collectively acting organisations. Thus, housing cooperatives shared with social democratic parties a common origin in the wider labour movement and common goals for “social” home-ownership (see chapter on housing). Furthermore, the social bases of both housing cooperatives and social democratic parties overlapped in the sense that the majority of inhabitants of cooperative dwellings voted for the social democratic parties and belonged to the lower-to-middle middle classes in both countries. However, the abandoning of cooperatives’ social-housing orientation and, in Sweden, an alteration of the voting behaviour of cooperative inhabitants since the 1990s, eventually weakened former congruencies between housing cooperatives and their inhabitants on one side and social democratic parties on the other side. Right-wing parties, on the other side, were less assertive in their policies towards housing cooperatives as social-collective organisations, but pursued a weakening of the social and collective set-up of cooperative housing. Instead, they altered legislation to enhance the individual ownership rights of inhabitants. • In sum, the support by social democratic parties for association- friendly policies towards housing cooperatives until the 1990s, and their more distanced stance afterwards, seems to support the hypothesis that partisan support for associations is based on, and varies with, interests in associations with congruent constituencies. Likewise, right-wing parties’ policies concerning housing cooperatives, which tended to weaken cooperatives as investors (in Sweden) and social-collective actors, could be interpreted as a lack of a special interest in the involved labour-movement-related associations. On the other side, it could be argued that social democratic parties only pursued an assertive policy towards associations as providers of “‘socially effective” housing. Housing cooperatives, indeed, originally tried to establish a “social” form of home-ownership, which was intended to be accessible to broader social strata. Against such an argument it could be said that cooperative housing is socially less accessible in comparison with public rented housing due to the demand for a capital investment. The necessity of making a capital investment was regarded by Swedish social democrats as a serious obstacle to those in the lower social strata, which was their argument to additionally opt for generous public rented housing. The Norwegian

191 Labour Party, on the other side, was more optimistic about the ability of those in the lower income strata to gain cooperative ownership. It could even be argued that other forms of associational rented housing would have been more in accordance with a socially inclusive rented housing strategy. However, social democratic governments did not support assisted forms of associational rented housing. They either turned semi-associational rented housing into publicly controlled rented housing, as in Sweden, or favoured cooperative ownership over all forms of rented housing, as in Norway. In sum, it could be argued that Swedish social democrats indeed pursued a policy mainly to support associations with congruent constituencies and not because they sincerely believed that cooperatives would be able to solve the housing problems they were concerned with. The policy of the Norwegian Labour Party, however, can also be interpreted as merely a strategy to solve the existing housing problems based on a welfare interest in socially inclusive home-ownership. Right-wing parties, on the other side, gave less support for housing cooperatives as social-collective organisations, but instead strengthened cooperative inhabitants’ individual ownership rights. This policy could be seen as being in accordance with their aim to further private market housing. Likewise, the support of Norwegian right-wing parties for subsidising cooperatives could be interpreted as a strategy to avoid public rented housing. Norwegian right-wing parties may have regarded housing cooperatives as the lesser evil in comparison with public rented housing, due to a lesser necessity, in general, to subsidise capital for cooperatives than for publicly rented housing. Swedish right-wing parties, on the other side, supported generous subvention of social rented housing at the time of its introduction, which, in comparison to housing cooperatives, requires a higher degree of capital subvention and a stronger administrative involvement of public agencies. This policy is difficult to account for as a general right-wing strategy to save tax revenues or to further the private housing market. In this regard, the policy of Swedish right wing parties could be interpreted as based mainly on a desire to avoid furthering labour-movement related associations.

192 Day-care associations In Norway until the 1990s, the main active day-care associations belonged to associations such as Christian missionary organisations, the Norwegian State Church151 and the Housewives’ Federation. These religious-conservative associations had in common that they defined their main organisational purpose as furthering Christian belief and/or as operating according to Christian principles. Religious- conservative associations provided predominantly short-time day-care for older children unlike, for example, parent cooperatives. Their social support bases and organisational strongholds were geographically situated in the Southwestern region of Norway, which is known for its high religiosity and, formerly, its strong lay movements. Like those associations, the Christian People’s Party originates in and gains its strongest voter support from the Southwest region of Norway. Furthermore, like those associations the Christian People’s Party was interested in religiously oriented day-care provision for older children for only several hours per week. In this sense, the • Christian People’s Party and religious-conservative day-care associations had overlapping day-care interests, overlapping ideological convictions and geo-culturally overlapping support bases. The Norwegian Labour Party, on the other side, traditionally had less voter support in the South western region of Norway. Partisan policies concerning those associations’ religiously oriented short-time day-care for older children diverged in the following way: The Christian People’s Party was the main driving force supporting statutory regulations allowing associations to select curricula, personnel and clients according to specific worldview associations, a policy for which they gained support from other right-wing parties. Furthermore, they pressured to prolong the generous subventions for short-time day-care for older children, for which they partly gained support from other right-wing parties. The Norwegian Labour Party, on the other side, was less supportive of the kind of day-care those religious-conservative associations provided. It lowered relative subsidies for short-time day-care and day-care for older children and attempted to delimit the operative independence of those associations. Thus, the diverging policies towards religious-conservative day-care associations, with the Christian People’s Party on one side and the Labour Party on the other side, could be regarded as based on

193 variations of congruent constituencies with the main active day-care associations in Norway until the 1990s. This seems to support the hypothesis that partisan policies vary according to interests in associations with congruent constituencies. On the other side, it could also be argued that the Christian People’s Party pursued an assertive policy towards religious- conservative day-care associations out of a right-wing interest to save tax revenues and to further closer-to-market associational operation. In addition, the Christian People’s Party’s welfare interests could be characterised as interests in effective conservative welfare provision, because they usually pursued a “middle way” between market liberal policies and social welfare policies. In this respect they could be classified as a Christian democratic party in a Continental European style rather than simply as a pro-market right-wing party. Religious-conservative associations certainly provided effectively a conservative form of welfare, but they could rather be said to have been relatively inefficient with respect to the need for rather high subvention levels for less cost-intensive day-care provision. By contrast, parent cooperatives were able to provide cost-intensive day- care with a relatively low degree of public subvention, but were not pursuing an “effective” conservative day-care, due to their care for younger children and long opening hours. Thus, the assertive policy of the Christian People’s Party seems not to have been guided by efficiency deliberations, but more by an interest in furthering effective conservative welfare provision by some specific day-care associations. Thus, their interest in religious- conservative day-care associations can be regarded as being in accordance with both their interests in congruent constituencies and their welfare interests in effective conservative welfare provision, but not with their interests in efficient welfare production. Since the 1990s, day-care associations in Sweden and Norway have included a wide range of small associations, of which parent cooperatives (and individuals) are the most prevalent organisation types. These are accompanied by less numerous providers, such as firms, Waldorf, Montessori and religious organisations, etc. These multifaceted associations can be assumed to derive from various social constituencies. The predominance of smaller owner categories of associations seems to reflect quite heterogeneous “micro-interests” ranging from pure practical interests to interests in parents’ self- organisation to pedagogical or religious interests. On the other side,

194 average day-care provision by associations is today rather similar to that of public providers with respect to opening hours and children’s age in both countries. To what extent can diverging partisan policies be understood as being based on congruent constituencies with those multifaceted associations? In Sweden, right-wing parties argued for their association-friendly policy152 efficiency gains and greater variation of associational provision in comparison with public provision. They asserted that associational provision would lead to reduced public spending and increased freedom of choice for parents. At the same time, they did not question municipalities’ control over associations, thus agreeing to a delimitation of the relative operational independence of associations. The only instance where they defended associational independent operation was in their opposition to maximum levels for parent fees. In tendency, Swedish right-wing parties’ policies towards associations seem to indicate an interest in tax savings, both from efficiency gains and from parents’ greater financial involvement in day-care. They tend to demonstrate to a lesser degree a special interest in the involved • associations comparable to the Christian People’s Party’s interest in religious-conservative associations or social democratic parties’ interest in cooperatives. The Swedish Social Democratic Party, on the other side, first restricted associations’ eligibility for statutory subsidies and introduced maximum levels for parent fees. The social democrats’ reluctance was motivated by their concern that subvention of associational provision would allow well-off parents to seek exclusive, higher-quality day-care outside the public system. Their introduction of maximum fees could be interpreted as reflecting this worry. In sum, social democratic disinterest could be interpreted as based on their interests in ensuring a socially inclusive day-care provision, which they worried would be lost with associational provision. In Norway since the 1990s, the only partisan conflict over associational provision occurred around the introduction of maximum levels for parent fees, which was, as in Sweden, decreed by left-wing parties against the opposition of right-wing parties. Right-wing parties feared that associations eventually would not be able to maintain their operations under the new conditions.

195 When interpreting these policies, one could, eventually, assume that the Christian People’s Party still defended the remaining religious- conservative associations, which may have been in need of extra parental payments. Otherwise, this recent conflict seems to indicate interests along a left-right pattern as in Sweden, whereby left-wing parties were interested in enhanced social inclusiveness and right- wing parties worried about increased public financial responsibility in the long term. In summary of these deliberations about partisan interests in associational provision, it is useful to distinguish between the period until the 1990s and the period afterwards. Until the 1990s, partisan interests in associations at least partly seem to indicate interests in specific associations, which shared congruent constituencies and/or the welfare interests of the supporting parties. In recent decades, however, partisan interests would be difficult to interpret as being based on congruent constituencies: partly due to a lack of predominant associations active in a welfare area, but partly also due to associations, such as parent cooperatives and today’s housing cooperatives, that lack the clearer socio-political attachment of former housing cooperatives and religious-conservative associations. When one would interpret newer partisan policies towards associations, only seldom is the assurance of associations’ rights at issue, if such rights would imply a delimitation of clients rights, such as former association rights to determine curricula disregarding parent wishes and housing cooperatives’ rights to impose collective sales regulations disregarding inhabitants economic interests. The only recent example was right-wing parties’ defence of associations’ rights to determine parent fees. Instead, policies either enhance clients’ rights towards the operating associations, of which enhanced ownership rights for cooperative inhabitants is an example, or policies regulate in detail associations’ operation, examples of which are the subjection of day- care associations to municipal day-care planning and the introduction of the maxtaxa. In sum, parties seem to be less interested in specific associations, but more in specific welfare production by any association. In conclusion, one could say that the hypothesis of partisan interest in associations, based on their interest in congruent constituencies, seems to have been applicable until the 1990s, but not afterwards.

196 In the following, the development from greater partisan interests in specific associations towards greater interests in welfare production is argued to be influenced by wider developments of organisational change in both countries, which affect associations, parties and their inter-relations alike.

4. Outlook: Changing parties and associations Associational welfare in housing and day-care underwent some changes in Sweden and Norway, which resulted in an alteration of the operation and/or composition of active associations and, furthermore, an alteration of their connections to political parties. In the following, it will be argued that these changes are based on altering individual “bottom-up” support for existing associations and political parties alike. In a first step, I will describe altered individual support patterns for associations and parties, leading eventually to an altered composition of main associations and parties in parliament. Afterwards, I will turn to the effects of altered individual interests on parties and associations’ inter-organisational connections in Sweden and Norway. Finally, some current theoretical considerations of these • developments will be briefly presented. Individual support for associations changed in both countries, leading eventually to an altered composition of associations. In the case of housing, the cooperative organisation persisted as such, but the actual relevance of cooperative supra-organisation was weakened in favour of individual cooperative owners’ influence. In Norway, in the 1980s, inhabitants of housing cooperatives publicly rebelled against the cooperative leadership in order to accomplish the abolishment of sales and price restrictions (Gulbrandsen 1983, 1988, Annaniassen 1996a, b). In Sweden, however, it is not so obvious whether the abolishment of cooperative restrictions came in response to pressure from inhabitants or if it is due to political interference, but today cooperative leaderships seem to assume that it would not be possible to reintroduce restrictions.153 Thus, one could speak of weakened loyalty to the original social and collective ideology of housing cooperatives, which even included an acceptance of a delimitation on individuals’ own say about their property. Instead, support for cooperatives became more instrumental, and individuals’ influence on their property was strengthened. In the case of day-care, religious-conservative-based associational organisations were diminished in favour of a large variety of

197 associations, namely parent cooperatives, small privately owned facilities, several pedagogical facilities, Churches, etc. The decline of religious-conservative associations in favour of new day-care associations has been accompanied by a lengthening of associations’ opening hours and a lowering of the age categories, which could be a reflection of rising interest in day-care to supervise children during the working hours of their mothers. Furthermore, the decline of religious- conservative associations, which were entitled to specify curricula, even against parents’ wishes, in favour of newer associations that do not have the same entitlement, led, in this sense, to a strengthening of individual influence. In sum, one could speak of a weakened ideological support for specifically religious day-care education for children and/or for the underlying conception that mothers should be housewives. Instead, interests in day-care supervision seem to be either of a more practical and economic nature or still of an ideological nature, but more diversified into various ideologies and less encompassing, such as today’s religious associations or Waldorf associations. Furthermore, individual influences on associational provision are supposed to have been strengthened, while the influence of associations over their own operations has been diminished. Individual support for associational welfare engagement seems to have shifted from associations that also offer collective-ideological guidance, to individual support for associations that address more individualistic, often practical, interests in welfare production. As a result of these changes in individual support for associational welfare engagement one could speak of changes in the composition of associations active in a specific area. The pattern of associations shifted from being dominated by few, encompassing associations pursuing specific welfare ideologies and relying on loyal constituencies, towards domination by either large associations, which act merely as welfare producers for clients with individual, practical interests, or by several diverse “micro” associations, some of which are still able to pursue ideologically shaped welfare production, but on a less encompassing level. In this sense, one could speak of an alteration and diversification of the associational composition in reaction to individualised, often de- ideologised, welfare interests and enhanced individual influence over associational provision.

198 The above-mentioned associational changes resemble changes that students of political party organisations call a development from a predominance of mass parties to catch-all parties, or from mass parties to electoral-professional parties (Kircheimer 1966, Panebianco 1988). According to research on political parties, parties are confronted with various changes. First, political parties in general saw a decrease in their core electorates and memberships, while confronted with an increase of electoral volatility (Aardal and Valen 1997, Holmberg and Oskarsson 2004, Mair, Plasser and Müller 2004a, b). This development indicates decreased loyalties of voters and citizens and confronts party elites with less predictable electoral markets. Party policies and electoral agendas became less ideologically “branded” and increasingly based on issue-assemblies, addressing a more heterogeneous electorate (Listhaug 1997, Knutsen 2001). In their quest for electorates, parties seem to target increasingly narrow groups of citizens – single parents, students, etc. – and try to aggregate these group appeals (Mair, Plasser and Müller 2004a). A development, which also Swedish and Norwegian parties were confronted with.154 • Second, party patterns showed an increased diversification. Older parties were confronted with a decline of vote shares or, eventually, ceased to exist, while new parties established themselves. As a result, the compositions of parliaments show an increase in the number of represented political parties (from 5 to 7 parties in Sweden and from 6 to 7 in Norway from 1945 to 1990s, Urwin 1997). The social democratic parties were confronted with similar changes in Sweden and Norway. Their memberships have declined since the 1980s and their core electorate of working class voters showed an ongoing decline (Svåsand, Strøm and Rasch 1997; Pierre and Widfeldt 1994). Their electorate is increasingly based on new middle-class voters, especially those who belong to lower and middle service classes, but nevertheless has declined from around 50% to around 35%. Furthermore, new middle-class voters tended to be a less stable pool of voters for the social democratic parties than workers were, which left social democratic parties with an overall more volatile electorate (Aardal and Valen 1997, Holmberg and Oskarsson 2004, Knutsen 2001; Evans 2004). The Christian Democratic Party, on the other side, saw a slighter decrease in their overall vote shares and core electorates. Religiously active voters remained a large group among their supporters (63% in

199 1965 to 52% in 1989), but other types of formerly loyal voters declined (for example, teetotaller voters declined from 84% to 56% between 1965 and 1989). Their share of female voters declined during the 1980s, but rose again since the middle of the 1990s. Membership numbers started to decline in the 1980s, but the decline has so far not been pronounced (Svåsand 1994). Thus, social democratic parties can no longer rely on the stable, class-based electorate of the past, which provided reliable support and served as a source of guidance for party ideology and policy positions. The Christian People’s Party, on the other side, can still be regarded as a party with a strong core electorate with a relatively homogeneous ideological orientation. In sum, both forms of organisation – parties and associations – saw a loss of loyal supporters, and were instead confronted with increasingly heterogeneous interests of members, clients or voters. This led to an organisational composition involving both ideologically “bloated” organisations with diversified constituencies, and smaller “micro” organisations with still rather homogeneous constituencies, that, to some extent, are pursuing ideological aims. In addition, students of voluntary organisations at large, such as sports clubs, hunting clubs, youth associations, political parties, unions, religious associations, art clubs, immigrant groups, environmental groups, and nature associations, describe developments similar to those observed here concerning parties and associations (Selle and Øymyr 1995, Engberg 1986, Klausen and Selle 1995a). According to these authors, the relevance of older, often ideological organisations declined over recent decades, while newly founded voluntary organisations became increasingly significant. In Sweden, membership in organisations such as the free churches and teetotaller organisations declined. On the other side, more leisure time and narrower-interest organisations such as village or community groups were founded after 1950 (Engberg 1986; Johansson 1980; Lundkvist 1977). Likewise, in Norway, several traditional organisations such as missions and temperance organisations declined, while musical, cultural and leisure organisations increased (Selle and Øymyr 1995, Tranvik and Selle 2007). These changes were accompanied by increased turn-over rates of memberships in organisations (Wollebaek and Selle 2002), or, in analogy with party- research terminology, they experienced an increase in “membership volatility.”

200 One could add that economic interest organisations also saw a decline in member loyalties, leading to a more diversified composition of main organisations in both countries. Union affiliation, for example, did not decline on a large scale, but became increasingly diversified in terms of a weakening membership share in the older unions belonging to the Norwegian and Swedish LO and increased membership in newer unions such as TCO and SACO in Sweden or YS, Akademikerne and Unio in Norway. The same could be said about housing interest organisations, where older organisations such as tenant (and cooperative) organisations lost their primacy to newer home-owner organisations with increasing memberships (Engberg 1986). The old consumer movement, on the other side, is marked by an increasingly practical attitude of members, who not only buy at or hold memberships in traditional consumer organisations such as Co-op, but increasingly buy at and hold “club” memberships in commercial retailers such as Ikea and H&M. In sum, individual support for various organisations seems to have undergone changes, which has affected various areas of social organisation. This led often to an altered composition of parties, • associations and other organisations. Altered individual support can also be argued to have led to a weakening of inter-organisational connections between various types of organisations. As has been described in this study, formerly close links between associations and political parties loosened, while newer associations have not built up connections to political parties. This development between parties and associations can be regarded as part of a wider dissolution of inter-organisational connections between parties, associations and other organisations. The older religious day-care associations and housing cooperatives, for example, had been part of wider networks of church and labour organisations within broader social movements. However, housing cooperatives dissolved connections, for example, to the tenant movement and to unions. Religious day-care associations, on the other side, did not weaken their ties to churches or lay organisations, but were replaced by newer associations. Newer day-care associations are seldom associated with ideologically like-minded organisations or social movements. Likewise, political parties either deliberately reduced connections to external organisations or saw former allies being diminished. Social democratic parties detached themselves from (blue-collar) unions

201 (e.g., the Swedish Socialdemocratic Party abolished collective membership of LO members in 1990, the Norwegian Labour Party did the same in 1992; Ebbinghaus 1993). The denominational Norwegian Christian Peoples’ Party, on the other side, was confronted with a decline of religious organisations and conservative women’s organisations (Selle and Øymyr 1995, Selle 1997, Urwin 1997). In sum, one could speak of changed individual interests leading partly to a recomposition of prevalent organisations, partly to a restructuring of surviving organisations and, overall, to a loss of former inter-organisational connections.

How can these changes be understood? When trying to estimate theoretically how the relations between social interests and party organisations changed, students often refer to Rokkan’s systematisation of party systems in the 1960s to contrast the changed situation in the 1990s (Bartolini and Mair 1990, Mair 1997). While Rokkan described party systems and party organisations as patterned by mass mobilisations around relatively broad economic and cultural interests, students today regard party organisation as far less structured by patterns of mass interests. Rokkan described party political systems as patterned along social cleavages, i.e. major economic and cultural interests and conflicts, such as the labour-capital cleavage and the state-church cleavage, which led to mass mobilisations with various “cleavage-specific organisations” (Rokkan et al. 2001). In the case of the labour-capital cleavage, for example, Rokkan mentions socialist parties, unions and consumer organisations as housing interest organisations representing the “labour” interests of this cleavage (Rokkan et al. 2001). In this sense, mobilisations and organisations were structured along cleavage lines. Rokkan considered, furthermore, that political parties that became representatives for such cleavages could in turn rely on relatively stable and large pools of core voters that remained loyal to the political representatives of their main interests. In Rokkanean terms, voter support became aligned along these conflicts. When considering the political party-welfare association constellations examined here, one could speak of a Rokkanean constellation of “cleavage-specific organisations” that were shaped by mass mobilisation of widespread ideological or social interests.

202 The Rokkanean constellations, observed in this study, could be summarised as follows: Rokkan argued that the labour-capital cleavage was predominant in Sweden and Norway.155 Typical for both countries were strong and highly mobilised and unified labour movements with high and stable voter support for the social democratic parties, comprising working- class and lower-middle-class voters. Furthermore, labour organisations (e.g., unions, consumer interest organisations and housing interest organisations) were able to gain large memberships and were usually organised in only a few major national organisations. A major difference between Norway and Sweden is, according to Rokkan, a cultural cleavage in Norway, which partly “cross-cut” the labour-capital cleavage (cp. Rokkan and Valen 1954; Rokkan and Hagvet 1987). Norway was, to a stronger degree than Sweden, influenced by religious lay movements156 and other cultural organisations (Rokkan and Hagvet 1987, Rokkan and Valen 1954).157 In addition, a Christian democratic party was formed early in Norway and could afterwards rely on about 10% of the votes (it was founded in 1933, while it was established in Sweden in 1964).158 • The religious-cultural movement was predominantly concentrated in the Southwest of Norway. Distinct from the rest of Norway, the population of the Southwest had stronger conservative and religious orientations, higher membership rates in religious organisations and higher voter support for the Christian People’s Party. In addition, it showed less support for economic interest organisations and less voter support for the Labour Party (Rokkan and Hagvet 1987, Rokkan and Valen 1954, Øidne 1957, Selle and Øymyr 1995).159 The constellation of welfare associations and political parties, which has been theme of this study, may be seen as being based on the same cleavage pattern. The associations and political parties examined here originated either in the labour-capital cleavage or in the cultural cleavage, articulated by either well-organised labour movement organisations or by well-organised lay organisations based in the culture of the Southwest of Norway. Respective parties and associations shared interests and ideologies, which allowed for a common pursuit of welfare measures and support for associational welfare establishment. Furthermore, they could rely on loyal support from their respective social bases, or constituencies, which supported those organisations by voting and by joining the specific welfare measures offered by associations.

203 The Rokkanean constellation remained stable over some decades, but began to dissolve in the 1970s. Parties and associations did partly survive as organisations, but had to adapt to new conditions. Today, political parties and associations can no longer rely on the support of a larger pool of ideologically loyal members or voters, but need to pursue their success in different ways. In addition, the welfare activities established by associations lost their attraction in their original set-up to larger client groups and became either diminished or reformed (see above). Relations between parties and associations have likewise changed and today one can speak of single parties and associations that pursue their goals independently rather than of party-association relations. In research, two types of explanations are discussed concerning changes of Rokkanean (political party) constellations. First, changes are regarded as being based on changed social structural preconditions. Examples of such altered social structural conditions include secularisation and the decrease of older classes – the working class, farmers and old middle classes – and the rise of new classes such as the new middle classes. These changes resulted in an altered composition of voters’ economic and ideological interests, leading to a decline in votes for older parties, a rise of new parties and changed party policies. Concerning welfare associations, one could similarly relate altered structural preconditions for the welfare measures offered by associations to changes in associational successes and operations. Norwegian day-care associations could be said to have been affected by a restructuring of family life (eventually based on long- term secularisation), leading to increased demand for long-time day- care. The rise of multifaceted day-care associations in both countries, on the other side, could be based on higher educational levels, which led to a diversification of educational and ideological interests in multifaceted forms of day-care. Housing associations could be supposed to have been affected by overall enhanced housing conditions and housing supply and by easier access to capital on liberated financial markets, which may have led to middle-class interests in pursuing their housing needs individually. Alternatively, one could also say that the former rise of the welfare state created a “Versorgungsklasse” within cooperatives, which then demanded that they yield individually upon their housing assets on the housing market.

204 A second explanation for observed developments is seen in a process of generally decreased long-term loyalties among voters for specific parties, or de-alignment, applying to both old and new party formations. Indications of de-alignment are seen in increased voter volatility instead of long-term voter loyalty and member support and, in response to these new support conditions, a change in party tactics to decrease ideological party policies and to increase combinatory targeting of narrow citizen groups with various new issues (Mair, Plasser and Müller 2004a; Urwin 1997). Concerning welfare associations, one could similarly speak of decreased individual loyalty to encompassing associations that collectively pursued wider ideological goals. The two processes, de-alignment and social restructuring, are partly regarded as socially interrelated because the new middle classes are considered “to lack a prima facie loyalty to any political party” (Mair, Plasser and Müller 2004a). Students have characterised the change in organisational patterns from the first half of the 20th century to the end of the 20th century as a shift from organisation in popular mass movements towards • organisation in a “self-referential” style (“Eigenorganisering,” Selle and Øymyr 1995), “leisure associationalism” (Tranvik and Selle 2007) or “organised individualism” (Engberg 1986).160 Tranvik and Selle attempt to theorise about the decline of popular movements and the rise of “leisure associationalism.” According to Tranvik and Selle, the rise of affluence and spare time led to a change of individual interests. Today, individuals seek to fulfil their wishes instantly and lack the ideological patience for pursuing long-term goals (Tranvik and Selle 2007). However, this assumption may be too narrow, shaped by a focus on leisure associations, making it difficult to account for the rise of, for example, new unions or parent cooperatives, which both could be said to display individual efforts to pursue their goals with great patience over a longer time span.

205

Appendices

206 •

Appendix A

Appendix A table A1: Volunteering in social services in Western European countries

country Volunteering in full-time equivalents (FTE) Volunteers (persons) Main activity areas of volunteers All social services Social Children - services related services In Share of Share of Shares of population (%) Country Country specifics according thousands civilian population commonalitie: to literature employment (%) Care for (%) visiting services, spare-time for elderly:shuttleservices, refugees; physically disabledpersons, homeless, mentallyand addicts, alcohol anddrug 207 F 16 0,7 0,3 - - Youth groups G 97 0,5 0,1 4,2 2,2

NL 89 1,2 0,6 6,1 6,1 Child care, child and youth groups UK 219 0,8 0,4 7,5 7,8 Child and youth groups, urban and inner city areas N 7 0,3 0,2 - - Crisis centres for children and battered women S 14 0,4 0,2 4,0 2,5 Crisis and call centres for children and battered women, personal care 1in full-time employment equivalents (FTE) Source: Salamon et al. 1999; Kendall and Knapp 1996; Finalyson 1994; Sivesind et al. 2002; Lundström and Wijkström 1997; Jeppsson- Grassman and Svedberg 1995; Gaskin and Smith 1995

Appendix A table A2:Revenues of associational service provision, by sources in shares (%), 1990ies Countries Fields

Fees, charges Public revenues Private giving

Average of 22 Health 34 55 11 COUNTRIES* Education 45 47 8

Social services 37 45 18

France Health 13 80 6

Education 18 72 10

Social services 36 58 6

Germany Health 6 94 0

Education 23 75 2

Social services 30 66 5

Great Britain Health 51 23 26

Education 31 64 5

Social services 35 26 40

Italy Health 27 72 2

Education 50 49 2

Social services 36 57 7

Sweden Health 12 87 1

Education 42 54 4

Social services 20 71 10

Norway Health 14 82 4

Education 46 50 4

Social services 32 66 2

* Austria, Belgium, Finland, France, Germany, Ireland, Netherlands, Spain, United Kingdom, Czech Republic, Hungary, Romania, Slovakia, Australia, Israel, Japan, United States, Argentina, Brazil, Colombia, Mexico, Peru Source: Salamon et al. 1999, Sivesind et al. 2002, Lundström and Wijkström 1997

208 •

Appendix A table A3: Coverage levels of associational voluntary social insurances in Europe* membership of economically active population and (of total population), shares in % S DK UK B F CH

Start year- 1895- -1950 1895- -1930 1911- -1915 1895- -1940 1885- -1930 1915- -1970 End year of intervention

Membership - 4 (2) 99(44) 15(7) 104(47) 26(11)a 75(35) 3(1) 31(13) 5(2) 33(17) 21(10) 155(74) rates sickness insurances

S DK NL B F CH 209

Start year- 1935- -1970 1910- -1970 1920- -1945 1910- -1940 1920- 1935- -1970 End year of intervention

Membership - 2(1) 55(27) 9(4) 34(17) 13(5) 5(NA) 4 30(NA) 1 NA 8(4) 18(9) rates unemployment insurances

*Introduction dates of statutory subsidised voluntary insurances (Alber 1987): Sickness insurances: Sweden 1891, Denmark 1892, United Kingdom 1911, Belgium 1894, France 1898, Switzerland 1911 Unemployment insurances: Sweden 1934, Denmark 1907, Netherlands 1916, Belgium 1903, France 1905, Switzerland 1924 Source: Flora and Alber 1995: 75, 77; a calculation based on: estimation by Beveridge and Deakin 2000 and Mitchell 1988

Appendix A table A4: European housing stocks, shares of associational and public rented housing, %*

Ownership and tenure 1918 1930 1950 - 1960 - 1970 - 1980 - 1990 - 2000

NL Assoc 6 12 29 40 35 Public 6 11 12

Housing stock 2,1 2,5 4,3 5,8 6,7 (millions)

G Assoc 3 9 8 7 9 Public 4 10 10 5 10

210 Housing stock 18,3 14,7 22,6 26,3 37,3 (millions)

S Assoc 4 11 16 17 16 Public 7 13 22 23 24

Housing stock 2,1 2,7 3,7 4,0 4,3 (millions)

N Assoc 11 13 19 19 14 Public - - - - 4

Housing stock 1,0 1,3 1,4 1,8 2 (millions)

Source: see continued table

•

Appendix A table A4 continued*

Ownership and tenure 1918 1930 1950 - 1960 - 1970 - 1980 - 1990 - 2000 F Assoc 2 5 4 8 24 Public 17 15

Housing stock 13,5 15,6 21,5 22,1 24,5 (millions)

UK Assoc 0,6 < 1 <1 < 1 1 1 2 3 5 Public 0,3 10 19 24 28 28 29 20 14

211 Housing stock 7,9 11,5 14,5 18,0 23,7 25,4 (millions)

Source: International: Department of Environment, heritage and local Government Ireland 2004; Boverket 2005; Boelhouwer and van der Hejden 1992; Power 1993; Balchin 1996a; ILO 1924 and 1930 estimations about the share of associations within the social/regulated housing sector according to: Netherlands: Priemus 1990, Boelhouwer and van der Heyden 1992; UK 1918: Office of the Deputy Prime Minister/ Holmans 1987 • S: Hedman and Tham 1993; Bostads-och byggnadsstatistiska årsböcker • UK: until 1960, 1970 and 1980 without Scotland (Scotland 1921-1960: 1,1-1,6 dwellings; British Historical Statistics, Mitchell 1988); Office of the Deputy Prime Minister/ Holmans 1987 • NL: CBS; VROM; Priemus 1990; Hakfoort 2002 • F: INSEE Housing Surveys; Union HLM; Ministrère de l’Equipment et du Logement; Blanc 1993; Power 1993 • G: 1945-1990 FRG (West) otherwise in “current” borders; Wohnungswirtschaftliche Jahrbucher der gemeinnützigen Wohnungsunternehmen; Statistisches Bundesamt

Appendix A table A5: Core laws and directives about public authorities and associations within subsidy systems Sweden 1917, 1946 Subsidised loans for local authorities and associations 1948 Requirement for subsidised loans: Associations controlled by local authorities, except cooperatives 1992 Abolished requirement for municipal control Norway 1919 Housing Act Subsidised loans foremost to municipalities 1946 Lov om den Norske Stats Subsidised loans foremost for cooperatives husbank 1982 (reformed) Loan-issuing for local authorities Netherlands 1901 Woningwet Support for municipalities and associations 1990 Volkshuisvesting in de Transformation of municipal housing into associative housing Jaren Negentig 212 Germany 1924 Hauszinssteuer Subsidised loans for municipal and associative rented housing 1950 Wohnungsbaugesetz Subsidised loans for municipal and associational rented housing Great Britain 1919 Housing and Town Subsidies for municipal and associational housing Planning Act 1946 Housing Act Subsidies for local authority rented housing 1974 Housing Act Support for Housing Associations

France 1928 Loucheur Act Compulsory establishment of public housing societies for every geographic area; Subsidised loans for HBM (Habitations à bon marché): Employer housing societies (Sociétés anonyms d’habitation à bon marché), Public authorities (Office publics d’habitation à bon marché), Cooperatives (Sociétés cooperatives d’habitation à bon marché) 1953 Loi Courant Housing tax on enterprises: 1% of salary costs for housing purposes Source: ILO 1930 Holmans 1987, Power 1993, Union National HLM (former HBM), Blumenroth 1975

•

Appendix A table A6: Housing policies in Western Europe 1920-2000

- 1945 1945 - 1950 - 1955 - 1960 - 1965 - 1970 - - 1975 -1980 -1985 -1990 - 1995 - 2000

Netherlands I II III IV V

Germany1 I II

213 III IV V

Sweden I II III IV V

Norway I II III IV V

Notes and source: see continued table

Appendix A table A6 continued

- 1945 1945 - 1950 - 1955 - 1960 - 1965 - 1970 - - 1975 -1980 -1985 -1990 - 1995 - 2000

France I II III IV V

United I

214 Kingdom II III IV V 1 1945-1990: West Germany I Emphasis on regulations concerning hygienic and security standards, slum clearances, extension of public control over land usage and rights to build for social purposes, time limited rent regulations and subsidy programs II High degree of government involvement, particularly in order to alleviate housing shortages: rent regulations, subsidized loans and interest rates on a historical high level, further extension of public control over land usage III Greater emphasis of housing quality: subsidies for renovations and rehabilitation IV Shortening of public subsidies; greater emphasis on problems of housing distribution and targeting specific groups V Promotion of owner-occupation: tax exemptions, convergences from rented housing (private and social), targeted subsidies Source: Boelhower and Hejden; ILO 1924, 1930; Power 1993; United Nations 1976; Balchin 1996a, Harloe 1995

Appendix A table A7: Sweden Day-care facilities by ownership Non-public facilities Public facilities All facilities a

% counts % counts counts

1943 95 324 5 18 342 1950 725 1960 38 442 62 728 1170 1970 3 113 97 4211 4324 1980 1 135 99 9349 9484 1990 13 1986 87 12844 14830 1995 19 2354 81 10331 12685

a the sum of all facilities is overestimated due to double counts of different forms of day-care organised within the same facility ( “daghem, fritidshem”) Source: SOU 1943:9, Ds S 1965:3; Prop. 1966 nr.54, SOS Statistiska meddelanden, SCB, Skolverket Barnomsorg, skola och vuxenbildning various years

Appendix A table A8: Swedish central state day-care expenditure, millions of Swedish crowns 1950 1963 1965 1966 1970 1974 1976 1990 2003 • Central 2.0 7.0 23.3 51.5 82 495 1245 12268 15491 state 1 block grants to municipalities not included Source: Statsverksproposition, Bil.7: Socialdepartementet; 1990, 2003: NOSOSCO

Appendix A table A9: Norway Day-care facilities by ownership Public Associations All (counts)

% % 1960 15 40 85 219 259 1975 53 468 47 415 883 1980 54 1370 46 1184 2553 1990 48 2737 52 1911 4648 2000 51 2982 49 2848 5830 Source: NSD; St.meld. 89 1961-1962, NOU 1972:39

Appendix A table A10: Norway day-care expenditure, millions of Norwegian crowns 1970 1972 1978 1981 1990 2003 Central state 6.5 3.9 213.9 363.8 1921 301 1 block grants to municipalities not included Source: NOSOSCO, various years; St.meld.nr.93 1980-81

215 Appendix A table A11: Operational requirements Sweden Norway

Personnel density 60ies >3** 1:5 (incl. administrative - personnel) 70ies/80ies >3 1:5 1:14-18 <3 1:4 1:7-9 90ies >3 1:5 1:14-18 <3 2:5 1:7-9

Pedagogically trained personnel* 1957 1975

Children group size 60ies/90ies >3 15 Not regulated <3 10-12

* required pedagogically trained personnel ** >3: children three year-old and older, <3: under three year old children Sources: St.meld.nr.93 1980-81, St.meld.nr.27 1999-2000, Knutsen 1990, Råd och anvisningar 1963, nr.163, Socialstyrelsen anser 1978, nr.2, Rostgaard and Fridberg 1998

Appendix A table A12: Professional day-care personnel, Norway and Sweden, shares in % 1985 1990 2000 Sweden 90 - 98 Norway 31 28 31 Sources: Knutsen 1990, NSD, Skolverket Barnomsorg, skola och vuxenbildning

Appendix A table A13: Opening hours and age of attending children in Norway and Sweden, share in % Sweden Norway Sweden Norway Full-time enrolled children, ages 1-5, Enrolled children in day-care, ages 1-5, share of cohorts share of cohorts 1976 20 5 26 8 1982 42 12 47 24 1991 56 28 59 42 2000 76 41 76 65 Sources: Blom-Hansen and Henneberg 1998; Rauch 2005

Appendix A table A14: Municipal funding of non-public facilities, Norway (in %) 1986 1990 1995 2000

Municipal subsidies with conditions 59 43 32 42 Municipal subsidies without conditions 20 17 12 17 Without municipal subsidies 18 35 56 41 No response 3 6 - -

Total (N) 1436 1912 3246 2848 Source: Nicolaysen 2001, NSD

216 Appendix A table A15: Day-care subsidies, Norway

Nordli II Brundtland II Willoch I Willoch II Brundtland II

Opening Age 1980 1982 1983 1985 1987 1989 hours

06-15 Above 1240 1310 1450-1620 1505-1680 1830 2270 3 Under 3 16-20 Above 2480 3690 3750 3890 4495 5580 3 Under 3 21-30 Above 4000 4150 9100 6450 3 Under 16425 7700 3 31-40 Above 4960 7380 7590 7880 9100 12270 3 Under 7440 13650 13700 14220 16425 19560 3 • 41- Above 10340 13680 3 Under 18660 22220 3 31-/06- Above *4 *6 *5 *5 *6 *6 151 3 31-/06- Under *6 *10 *9 *9 *10 *10 151 3 1 average subsidy for day-care places with opening hours of 0-8/10/16 Source: Ot.prp.nr.23 1974-75; St.prp.nr.1 (1980-81, 1981-82, 1982-83, 1990-91, 1996-97), St.meld.nr.8 1987-88; Information from the Children and Family Department; Departmentsrundskrift 2004

217 Appendix A table A15: continued

Brundtland III Bondevik I3 Bondevik II

1991-2 -1997 1998 Extra subsidies to start-up and to convert Kontantstötte into day-care for younger children with long opening hours Opening Age 1996 Opening Age 1998-1999 2005 hours hours public private 0-8 Above 3 3920 6540 7330 Under 3 7200 13320 14950 06-15 Above 3880 9-16 Above 3 7850 13060 14670 3 Under Under 3 14400 26670 29930 3 16-20 Above 7210 17-24 Above 3 11770 19610 21990 3 Under Under 3 21600 39960 44860 3 21-30 Above 8820 25-32 Above 3 15690 26110 29300 3 Under 13420 Under 3 28800 53280 59810 3 31-40 Above 15790 33-40 Above 3 16260 27060 30380 3 Under 28010 Under 3 30000 55960 62830 3 41- Above 19630 41- Above 3 19630 34290 38500 3 Under 32670 Under 3 36140 67550 75820 3 31-/06- Above *5 41-/0-161 Above 3 *3 *4 *4 151 3 31-/06- Under *9 41-/0-161 Under 3 *5 *3 *3 151 3 1 average subsidy for day-care places with opening hours of 0-8/10/16 2 schoolreform: lowering of school age to 6 3 Kontantstötte,: home care benefit, which offers all parents with children aged 1-3 years who do not attend state-sponsored day-care the same amount of money as the state subsidy per place. Ot.prp.nr.23 1974-75; St.prp.nr.1 (1980-81, 1981-82, 1982-83, 1990-91, 1996-97), St.meld.nr.8 1987-88; Information from the Children and Family Department; Departmentsrundskrift 2004

218

Appendix A table A16: Sweden Day-care facilities by ownership, shares in %

A a C E HC O For-profit Public Total

1943 37.4% 6.7% 7.6% 15.2% 24.9% 6.7% 342 1960 28.6% 2.3% 6.8% 62% 1170

a Red Cross, Salvation Army, Swedish Housewives’ Federation, Frederika-Bremer- Association, Fröbelförbundet, etc. (Svenska Fattigvårds- och barnavårdsförbundet) A (Association), C (Church), E (Employer), HC (Housing cooperative), O (Others) Source: SOU 1943:9, Ds S 1965:3, Prop. 1966:54

Appendix A table A17: Sweden Day-care facilities by ownership, shares in %

O** PC EC A E For-profit Public Total

1970 2 98 4324

1980 99 9628

1990 11 2 87 12850

1995 1 8 1 2 0.1 2 81 14830 • 2000 9 *8 2 3 78 12685

*The share of enrolled children usually close to the share of institutions. In the case of parent cooperatives, though, the difference between share of institutions and share of enrolled children is higher than for other providers. The share of enrolled children in parent cooperatives was with 7% still highest of all non-public day-care providers in 2000 (see appendix A table A18). ** Montessori, Waldorf, Swedish State Church, etc. PC (Parent cooperative), EC (Employer Cooperative), A (Association), E (Employer), O (Others Source: SOS Statistiska meddelanden, SCB, Skolverket Barnomsorg, skola och vuxenbildning various years

Appendix A table A18: Sweden: Children enrolled in day-care facilities by ownership, shares in %

Public Parent Employee For-profit Others* All cooperative cooperative enrolled children 2000 85% 7%* 2% 4% 3% 314 894

* Montessori, Waldorf, Swedish State Church, etc. Source: Skolverket Barnomsorg, skola och vuxenbildning

219 Appendix A table A19: Norway: Day-care facilities by ownership, shares in % Public PO** P E PC I M* Facilities, total 1960 15 29 13 7 6 NA 30 259 1973 51 28 8 6 3 NA 3 422 1975 53 23 7 6 4 NA 6 883 1980 54 21 9 3 8 NA 5 2553 1984 58 19 9 2 10 NA 3 3148 1988 61 18 7 1 13 NA NA 4043 1990 59 23 6 1 3 3 5 4648 1996 48 3 6 3 16 17 8 6407 2000 51 3 5 3 14 15 9 5830 *others, housing cooperatives, foundations, joint-stock companies ** in 1990: including housing cooperatives; since 1996: Housewives’ Federation and Sanitary Association, pedagogical and ideological organisations PO (Private organisation), P (Parish, denominational organisation), E (Employer), PC (Parent cooperative), I (Individual), M (Miscellaneous) Source: NSD; St. meld. 89 1961-1962, NOU 1972:39

Appendic A table A20: Major shifts of official day-care provider categories Main ‘standing’ Employer Parent Parish Private Other categories: cooperative organisation Changes affecting main categories 1975-84 e.g. Denomination s, Individuals, Housewives, Womens’ 1986-88 Incl. housing New: Sanitary NA cooperatives Parish Association and pedagogical, denomin ideological ation organisations, etc. 1989-91 New: New: Housewives’ Individual Federation, Womens’ Sanitary Association, etc. 1996- New: pedagogical, ideational organisation Further categories, which eventually appear in public statistics: Foundations, joint-stock companies, housing cooperatives, etc. Source: NSD

220 Appendix A table A21: Governments in Norway Government Formation Government Government Parliament year parties majority majority Gerhardsen I 45 DNA Yes Left Gerhardsen II 49 DNA Yes Left Torp I 51 DNA Yes Left Torp II 53 DNA Yes Left Gerhardsen III 55 DNA Yes Left Gerhardsen IV 57 DNA Yes Left Gerhardsen V 61 DNA No - Lyng 63 H, KRF, V, SP No - Gerhardsen VI 63 DNA No - Borten I 65 SP, H, V, KRF Yes Right Borten II 69 SP, H, V, KRF Yes Right Bratteli I 71 DNA No Left Korvald 72 KRF, SP, V No Right Bratteli II 73 DNA No Left Nordli I 76 DNA No Left Nordli II 77 DNA No Left Brundtland I 81 DNA No Right Willoch I 81 H No Right Willoch II 83 H, KRF, SP Yes Right Willoch III 85 H, KRF, SP Yes Right Brundtland II 86 DNA No Right • Syses 89 H, KRF, SP Yes Right Brundtland III 90 DNA No Right Jagland 96 DNA No Right Bondevik I 97 KRF, SP, V No Right Stoltenberg I 00 DNA No Right Bondevik II 01 H, KRF, V No Right Stoltenberg II 05 DNA, SP, SV No Right Party classification: ’right’: Conservatives - Høyre (H), Christian People’s Party - KristeligFolkpartiet (KRF), Center - SenterPartiet (BP), Liberal - Venstre (V), Progress - ALangePartiet/FremskridtsPartiet (PP) ’left’: Labour - Det NorskeArbeiderparti (DNA), Left - SosialistiskVenstre (SV), NorgesKommunistiskeparti, SosialistiskFolkeparti Sources: Lindblad et.al. 1984, von Beyme 1982; Regjering.no, Stortinget

221 Appendix A table A22: Governments in Sweden Government Formation Government parties Government Parliament year majority majority Hansson III 39 S, C, FP, H Yes Hansson IV 45 S Left Erlander I 46 S Left Erlander II 48 S No Left Erlander III 52 S, C Yes Erlander IV 58 S No Left Erlander V 68 S Yes Left Palme I 69 S Yes Left Palme II 70 S No Left Palme III 73 S No Left Fälldin I 76 C, FP, M Yes Right Ullsten 78 FP No Right Fälldin II 79 C, M, FP Yes Right Fälldin III 81 C, FP No Right Palme IV 82 S No Left Palme V 85 S No Left Carlsson I 86 S No Left Carlsson II 88 S No Left Bildt 91 M, FP, KD, C No Right Carlsson III 94 S No Left Persson I 96 S No Left Persson II 98 S No Left Persson III 02 S No Left Reinfeldt 06 M, C, FP, KD Yes Right Party classification: ’right’: Center - Centerpartiet (C, BP), Conservative - Moderatera samlingspartiet (M, H), Liberal - FolkPartiet (FP), Christian Democratic - KristDemokratiskasamlingspartiet (KD), Sverigedemokraterna (PP); ’left’: Socialdemocratic – SAP-Socialdemokraterna (S), Green - MiljöPartiet (MP), Left - Vänsterpartiet-kommunisterna (V) Sources: Lindblad et.al. 1984, von Beyme 1982; Regeringen, Valmyndigheten

Appendix A table A23: Housewives’ Federation: Regional membership in shares of population, percentages (shares within Housewives’ Federation membership),

1935 1950 1970 1980 1990 2000

Oslo area 0.6 1.3 1.4 1.1 0.8 0.3 (31) (19) (22) (23) (24) (25) Southwest 0.5 2.1 2 1.3 1 0.5 (41) (48) (43) (44) (48) (53) Rest of Norway 0.2 0.9 0.9 0.4 0.4 0.1 (28) (33) (35) (22) (28) (22) National average 0.4 1.3 1.3 0.9 0.7 0.3 Total 9817 43752 49854 38599 27740 12525 Source: NOS Population and Housing Censuses, Housewives’ Federation

222 Appendix A table A24: Associational provision and Christian People’s Party, 1980 Municipalities with… Associational day-care facilities of all facilities, shares, % 2 None Low High % of Municipalities municipalities (N) KRF1 Lowest 27 14 24 100 104 Low 22 41 24 100 109 High 26 23 26 100 108 Highest 24 23 26 100 103 Municipalities 279 71 74 100 424 (N) Missing 30 1 KRF (Christian People’s Party) vote share in a municipality: highest = 20.5%-, high =11.2%-, low =3.45%-, lowest = - 3.45% 2 other associational day-care facilities as a share of all facilities in a municipality: high = 32.5% -, low = - 32.5%, none = 0 Source: NSD

Appendix A table A25: Households privately renting their dwellings, shares in % 1920 1945 1970 1990 2001 Norway 47 34 22 23 Oslo 95* 49 24 29 Sweden 51 30 20 Stockholm 64 20 * 1920 boarders of the municipality Oslo • Source: SOS/NOS Population and Housing Census, various years

Appendix A table A26: Sweden: Housing conditions Overcrowed according Overcrowed according Without Without Norm 11 Norm 21 toilet bath 1945 30 51 74 79 1960 13 43 1970 3 33 10 22 1975 1 10 1990 3 1 norm 1: 2 persons per room without kitchen in a dwelling; norm 2: 2 persons per room excluding kitchen and one other room Source: Heady 1978; NBO 1990

Appendix A table A27: Norway: housing conditions Overcrowed according Overcrowed according Without Without Norm 11 Norm 21 toilet bath 1946 16 36 71 88 1960 1970 2 12 28 34 1985 1 norm 1: 2 persons per room without kitchen in a dwelling; norm 2: 2 persons per room excluding kitchen and one other room Source: NBO 1990

223 Appendix A table A28: Sweden: New construction Central state subsidies Public utility companies,* for newly built dwellings, share of all newly built coverage of all dwellings, % dwellings, % Of which (circa) Associations (foundations, cooperatives, shareholder companies, employers) 1917-1923 High 45% 40 22 1924-1940 Low 15% 16 12 1941-1945 High 70% 33 27

* state, municipalities, associations Source: SOS Population and Housing Census 1920, 1933; ILO 1930; Silk 1948; SOU 1945: 63

Appendix A table A29: Sweden: Ownership of housing stock 1945-1990 (percentages) 1945* 1960 1970 1980 1990 State, municipal, etc. 7 13 21 22 23 Cooperative 4 11 14 16 17 Person, 89 64 60 firm, etc. of which privately rented 41 21 owner-occupied 36 41 not specified 2 2 Total 100 100 100 100 100 * not comparable with other years due to double counting (SOS Population and Housing Census 1960) Source: SOS Population and Housing Censuses

Appendix A table A30: Sweden: Tenure of housing stock 1945-2000 (percentages) 1945 1960 1970 1980 1990 2000 Public utility rented 6 14 23 24 25 23 Cooperative tenure 4 9 13 14 15 17 Privately rented 52 43 30 21 20 21 Owner-occupied 38 34 34 41 40 39

Total 100 100 100 100 100 100 Source: SOS Population and Housing Censuses, calculations for 2000 according to Berger (both cited in Bengtsson 2006)

Appendix A table A31: State Housing Bank regulations, loan coverage of project costs 1946 1982 1. Cooperative building societies, 90 1. Cooperative building societies, the 70-75 the statues of which are approved statutes of which are approved by the by the Housing Directorate Housing Directorate 2. Persons who are themselves 85 2. Municipalities, counties, state going to occupy the house 3. Other houses 75 3. Persons who are themselves 55-60 going to occupy the house 4. Other houses

224 120000

100000

80000

60000

completed dwellings completed 40000

20000

• detached multi-dw elling all

Figure A1: Sweden: Completed dwellings in all houses, counts Source: SOS Byggnadsverksamheten, SOS Bostads - och byggnadsstatistisk årsbok, various years

100%

80%

60%

40%

shares % in 20%

0% lling dwellings, dwellings, detachedlling and multi-dwe

detached multi-dw elling

Figure A2: Sweden: Completed dwellings in detached and multi-dwelling houses, shares in % Source: SOS Byggnadsverksamheten, SOS Bostads - och byggnadsstatistisk årsbok, various years

225 Appendix A table A32: Cooperatives’ ownership, affiliation and independence, shares, %, calculated*

1945 1960 1970 1980 1990 20001 S Affiliated 45 52 64 68 64 HSB 37 35 46 44 47 44 Riksbyggen 8 17 20 21 20 Independent, Others2 55 48 36 32 36 Total 90377 299007 458230 584414 683000 736000 N Affiliated (NBBL) 69 73 64 84 Independent 31 27 36 16 Total 168563 258570 336807 277000 1 Sweden 1997 2 e.g., SBC, SKB * calculations based on cooperatives’ own statistics and official statistics, which are not

70% y, % y, 60%

50%

40%

30%

20%

10%

0% share of completed dwellings by investor categor investor by dwellings completed of share 1949 1954 1959 1964 1969 1974 1979 1984 1989 1994 1999 2004

public private affiliated independent Figure A3: Sweden: Completed dwellings in multi-dwelling houses by investor category, shares in % Source: SOS Byggnadsverksamheten, SOS Bostads - och byggnadsstatistisk årsbok, various years

226 80%

ory, % ory, 70%

60%

50%

40%

30%

20%

10%

0% share of completed dwellings by investor categ dwellings investor by completed of share 1953 1962 1971 1980 1989 1998 private affiliated cooperative independent cooperative, joint-stock company public other

Figure A4: Norway: Completed dwellings by investor category, shares in % Source: SOS Byggnadsverksamheten, SOS Bostads - och byggnadsstatistisk årsbok, various years • Appendix A table A33: Norway: Subsidised dwellings by investor category, shares in %* private public Affiliated cooperative Independent cooperatives, joint- stock company 1969-81 78 0 88 49 1982-98 65 100 90 9 * calculated. Should be interpreted carefully, due to insecure new construction data. Sources: State Housing Bank, NOS Byggearealstatistikk

Appendix A table A34: Norway: Households by tenure and socio-economic status, percent, 1995 All Cooperative Owners Renting households tenants tenants, others

Farmers, fishers, foresters, etc. 2 0 3 1 Self-employed 7 4 9 4 Higher salaried 10 9 13 6 Middle and lower salaried 28 40 31 26 Workers 17 15 20 19 Others (other economically 25 31 23 44 active, economically inactive) All 100 100 100 100 n 4455 578 2493 979 Source: NOS Survey of Housing Conditions 1995

227 Appendix A table A35: Norway: Household income by housing situation 1973 1994 Cooperatives, all Cooperatives, all shareholders shareholders Lowest1 5 7 Lowest2 5 3 Low 9 17 Low 17 8 Middle low 11 18 Middle low 16 14 Middle 18 20 Middle high 22 14 Middle high 20 16 High 18 8 High 14 23 Highest 9 15 Highest 9 36 Total 100 100 100 100 n 378 2385 Total 578 4455 1lowest: under 10 000 nkr, low 10 000-19 000, middle low 20 000-29 900, middle 30 000-39 900, middle high 40 000-49 900, high 50 000-59 900, highest 60 000 and over. 2lowest: under 60 000 nkr, low 60 000-99 999, middle low 100 000-179 999, middle high 180 000-259 999, high 260 000-379 999, highest 380 000 and over. Source: Survey of housing conditions 1973, 1995, NOS

Appendix A table A36: Sweden: Percentage-differences, national elections, %-%, 1976 SV SAP C FP M N

Multi-dwelling (Coop-)1 -1 9 -8 3 -2 254 owned Rented 2 7 -6 -1 2 869

Detached Owned -2 -7 6 1 3 1159 Rented -1 3 4 -5 -2 144 Others2 1 -16 14 0 0 100 Total 4 44 22 11 13 2526 1Single dwellings in Sweden cannot be owned individually 2 old-age housing, facility housing, etc. *SV (Left Party), SAP (Socialdemocratic Party), C (Centre Party), FP (Liberal Party), M (Conservative Party) Source: Swedish Election Studies, SSD

Appendix A table A38: Sweden: Percentage-differences, national elections, %-%, 2002 SV SAP C FP M N

Multi-dwelling (Coop-)1 1 0 -2 6 1 317 owned Rented 3 1 -3 -2 -3 597

Detached Owned -1 1 2 -1 1 1248 Rented 1 -4 2 -7 1 84 Other2 6 -6 -2 -3 -8 36 Total 8 34 5 14 11 2282 1Single dwellings in Sweden cannot be owned individually 2 old-age housing, facility housing, etc. *SV (Left Party), SAP (Socialdemocratic Party), C (Centre Party), FP (Liberal Party), M (Conservative Party) Source: SSD Swedish Election Studies

228 14 60 13

54 53 12 50 48

10 43 9 40 8 37 8 34

6 5 percentages, % percentages, 20 4 percentagedifferences, %-% percentagedifferences,

10 2 1

0 0 0 1976 1979 1985 1994 1998 2002 coop socialdemocratic voting minus average socialdemocratic voting coop socialdemocratic voting

Figure A5: Sweden: Socialdemocratic voting by cooperative owners, shares in % • Source: SSD Swedish Election Studies

Appendix B In both countries, interests in an increase of female labour market participation had influenced decisions about statutory intervention into day-care (Leira 1992, Kyle 1979). The alternative option for an expansion of labour supply could have been immigration of foreign employees, but was not considered of the Swedish and Norwegian governments as the only option (Leira 1992, Kyle 1979, Lundh and Olsson 1994). Immigration occurred, but never reached the same level and importance for labour supply in Sweden and Norway as, e.g., in Germany( see table A40).161

Appendix B table A40: Netto population increase rates 1965 1967 1968 1969 1970 1971 1972 1973 Averages 1965-73 S 4,4 1,3 1,6 5,5 6,1 0,4 -1,5 -1,4 2,2 N -0,5 0,8 0,8 0,3 -0,3 1,8 1 0,8 0,5 G 5,8 -3 4,6 9,4 9,3 7,6 5,4 6,2 5,3 Source: Mitchell 2003; OECD Historical statistics; OECD Labour force statistics

229 In Sweden, the first introduction of national subsidies occurred in the 1940ies. During WW II, Sweden experienced rising economic activity and labour force demand rose. The labour market situation was further acerbated due to war related mobilisation (Silenstam 1970; Tuveng 1946). In 1943, the state, the biggest trade union (LO) and the central employer organisation (SAF) agreed in subsidising day-care on a national level (SOU 1943: 9; SU utlåtande nr. 260 1943). The Labour Market Commission162 (Arbetsmarknadskommissionen) and several regional Labour Boards (Länsarbetsnämnd) recommended subsidies for day-care with the argumentation, that there were severe labour shortages and that day-care would make labour market access for mothers easier (SOU 1943:9; 10; Remissyttrandet nr. A 3803/42)163. Expansion of day-care subventions was initiated in the 60ies. It followed the same pattern as in the 40ies. Sweden had already in the 60ies labour shortages. From 1963 on, day-care subventions were stepwise increased (SFS 1963: 379, SFS 1966: 173, SFS 1973: 1205, 1207). Again, economical and political interests groups were pressing for an enlargement in publicly provided day-care. The unions (LO and TCO),164 the central Swedish Employers Organisation (SAF) and several labour market commissions had repeatedly demanded an extension of day-care (SU 1962:5165; Ds S 1963:3, see also public comments on proposition of the largest unions, LO and TCO, and the Swedish Employer Organisation to Familjeberedningen 1968; Kyle 1979). Appendix B table A41: Real GDP Per Capita Growth rates (year to year averages)

1950-60 60-70 70-73 73-79 Sweden 2,9 3,6 2 1,5 Norway 2,7 4 4,1 4,1 OECD 3,9 2 Source: Lane, McKay, Newton 1997; OECD Historical Statistics

Appendix B table A42: Official unemployment rates

1950 1955 1960 1965 1970 19751 1980

Sweden 2,2 1,7 1,4 1,1 1,4 1,6 2 Norway 2,7 2,5 2,5 1,8 0,8 2,3 1,6

1 rates are comparable from 1950 to 1970 and from 1975 until 1990 Source: Mitchell 2003; OECD Historical statistics; OECD Labour force statistics

230 In Norway, day-care expansion came with the Day-care Act in 1975. After several years of economic growth and low unemployment rates (see tables A41 and A42), the situation on the labour market tightened further at the beginning of the 70ies in Norway. Oil extraction from the North Sea, the lowering of retirement age in 1972 and the experience of sinking national labour market participation rates led to expectations of exacerbated labour shortages. The government expected a sharp rise in labour demand (Bull 1979; NOU 1972:39, Leira 1992).166 The Ministry of Government Administration and Consumer Affairs reasoned that an increase of the labour force necessitated an increase of day-care supply, which would enable mothers to enter the labour market (NOU 1972:39, Ot.prp.nr.23 1974-75, Leira 1992). When the economic situation of high growth rates and low unemployment changed since the 1980ies (see tables A43 and A44), day-care expansion, however, became not revised, but further extended. This may be due to entrenched female participation on the labour market and female representation in unions and parliaments (see table A45). • Appendix B table A43: Real GDP Per Capita Growth rates (year to year averages) 70-73 73-79 79-89 89-00 1970-2000 Sweden 2 1,5 2 1,3 1,6 Norway 4,1 4,1 2,3 2,7 3 OECD 3,9 2 2 1,8 2,1

Source: Lane, McKay, Newton 1997; OECD Historical Statistics

Appendix B table A44: Official unemployment rates

1970 19751 1980 1985 1990 1995 Sweden 1,4 1,6 2 2,8 1,7 7,7 Norway 0,8 2,3 1,6 2,6 5,2 4,9 1 rates are comparable from 1950 to 1970 and from 1975 until 1995 Source: Mitchell 2003; OECD historical statistics; OECD Labour force statistics

231 60

20 (respectively), % (respectively),

10 labour forcelabour as share of population 0 1930 1940 1946 1950 1960 female labour force Norw ay female labour force Sw eden labour force Sw eden labour force Norw ay

Figure A6: Female labour force participation 1930 – 1960, Sweden and Norway Source: SSB, SCB, Silenstam 1970

30 % 20

0 1960 1970 1980 1990 Sw eden Norw ay female labour forcelabour female as shareforce, labour total of

Figure A7: Female labour force participation 1960 – 1990, Sweden and Norway Source: Lane, McKay, Newton 1997

232

Appendix B table A45: Female representation Female Main union federation female representation in parliament, membership, (LO) % % N S N S 1939 1945 12,3 16,9 1950 17,1 19,0 4,6 9,5 1960 18,0 22,9 6,6 13,4 1970 23,5 30,0 9,3 14,0 1975 26,6 36,4 15,5 21,4 1980 32,5 43,2 23,8 26,4 1990 40,8 47,8 35,8 38,1 Sources: Lane et al. 1997, Ebbinghaus and Visser 2000

Appendix C Historically, statutory intervention into ‘housing’ was an intervention into an already existing private housing market. Households already spend large amounts of their income for housing consumption and mass housing was provided for before the state intervened (Harloe 1995). • The effect of housing policy on housing supply and demand is discussed by scholars. McCrone and Stephens estimate that “in all countries, regardless of the average standard of living, there is a large section of the population that cannot afford the full economic cost of what would generally be regarded as an adequate or tolerable standard of housing.[…]In most European countries, in the absence of such intervention, something of the order of a quarter to a third of the total population would be unable to pay the full economic cost of the housing it occupies” (McCrone and Stephens 1995). Some studies indicate that housing policies can influence the efficiency of housing production and thus enhance the quality and quantity of available housing (Barlow and Duncan 1994). However, exact assessments of the influence of housing policy on housing markets are difficult. Furthermore, housing policies are often subordinated other political concerns as for example overall macro-economic strategies for national economies, which affect for example costs of housing production via interest rate level changes or the regulation of financial markets. For example, the strong intervention of states into the housing market after the Second World War was not only pursued due to housing objectives as housing standards and supply, but likewise out of Keynesian considerations for the overall economic

233 development. Another example was the deregulations of financial markets during the 1980ies, which also affected mortgage lending and housing markets, but were not primarily introduced for housing purposes. However, Boelhouwer and Heijden regard that “it is also clear that there is some scope for the pursuit of purely political objectives” (Boelhouwer and Heijden 1992). States nevertheless try to influence housing markets and there is today no western European country without nationally subsidised housing (Balchin 1996a). States influenced housing markets through a variety of measures, which Oxley categorised as either subject subsidies for consumers or object subsidies targeted at the cost of dwellings (Oxley 1984). Lundqvist summarised possible courses of state interventions in a similar way, distinguishing production and consummation of housing (see figure A8, Lundqvist 1992a). States can intervene into the housing market by regulating central input factors as land, labour or capital. Land regulations are for example obligatory area plans, prescribing the use of defined areas for housing construction. Capital markets are for example influenced by the establishment of financial institutions or regulation of financial institutions’ functioning, but also by regulating the production of dwellings after quantity and quality criteria. Further regulations can be applied to influence the quantity and quality of dwelling construction (such prescriptions were often made in combination with land development planning). Regulations can also be made concerning the matching of individuals and dwellings through allocation and distribution prescriptions, or the specific forms of available ownership “titles” and the specific rights and duties connected to forms of ownership. Governments can also intervene through subsidising housing production or consumption. The subvention of costs for housing production can for example take the form of reduced VAT for production materials. The price of dwellings can furthermore be influenced through capital costs financing. Concerning housing consumption, states may grant for example subject subsidies as housing allowances or reduce income taxation for capital costs payments.

234 A third possible way of government action can be the direct ownership of either input factors for housing production, e.g. land, or ownership of dwellings.

•

Figure A8: Possible state interventions into the housing sector Source: Lundqvist 1992a

235 State intervention focused over time differently on types of measures in North-Western Europe (Boelhouwer and Heijden 1992, Danermark and Elander 1994a, Balchin 1996a, etc.). Following international comparative housing research, one could roughly distinguish three main phases of statutory intervention: first, a pre-phase of predominantly regulating housing markets until the Second World War, a second phase until the 1970ies with an extension of regulations and massive financial support through object and subject subsidies and a third phase with reduced financial support and a shift from object subsidies to subject subsidies. Reasons for shifts of policy are discussed out of several theoretical perspectives, each emphasising different factors. However, scholars roughly agree upon, that historical courses of state intervention are (also) based on variations of socio-economic and demographic preconditions: from times of high urbanisation levels, high population growth levels and inappropriate housing conditions in the decades after the second world war to a time with decreased growth rates, a higher quality of the existing housing stock and increased demands of wealthier consumers. Furthermore, a shift from Keynesian to neo- classical economic strategies and technical developments are sometimes regarded as an underlying reason for shifts in housing policies towards housing types. Besides such relatively common trends of international housing policies, more marked differences of countries’ housing policies exist concerning ownership and tenure types. Most scholars agree upon that such differences are to a stronger degree influenced by political factors and objectives (e.g. Boelhouwer and Heijden 1992, Lundqvist 1992a, Harloe 1995, Balchin 1996a, Danermark and Elander 1994a, Kemeny, McCrone and Stephens 1995; but not socio-economic functionalists as Donnison and Ungerson 1982; see table A46). Limited-profit housing is most prevalent in the Netherlands, the United Kingdom and Germany, where social housing was favourably subsidised (Tomann 1996, Boelhouwer and Heijden 1992, Holmans 1987).On the other side, private landlord housing has still a distinct share of the housing market in Germany and Switzerland, where private landlords achieved favourable statutory subvention. In Great Britain, on the other side, where private landlords were precluded from statutory subvention, it nearly disappeared (Tomann 1996, Lawrence 1996, Holmans 1987).

236 Appendix C table A46: Estimated gross domestic product per capita and housing tenure, western Europe, 1994-5 GDP $ per Owner- Private Social Other capita occupation rental rental tenure Private rented sector above the EU average: Switzerland 36,430 31 60 3 6 Germany 26,000 38 36 26 - Luxembourg - 67 31 2 - Belgium 22,260 62 30 7 - Social rented sector above the EU average: Denmark 29,010 50 24 18 8 Austria 25,010 41 22 23 14 France 23,550 54 21 17 8 Sweden 23,270 43 16 22 19 Netherlands 21,300 47 17 36 - Owner-occupation above the EU average: Norway 26,590 60 18 4 18 Finland 20,410 72 11 14 3 United Kingdom 18,950 66 10 24 - Italy 18,400 67 8 6 19 Spain 12,500 76 16 2 - Greece 8,400 70 0 4 Note: data give no specifications about ”other tenure” and limited information about ownership • Source: Balchin 1996a, CECODHAS (European Liaison Committee for Social Housing) 1995

237 Notes

1 For a definition of formal organisation, see Scott 2004 and Weber 1972: Verband, Verein, Betrieb. 2 In Weber’s terms, I refer to the distinction of communities Vergemeinschaftung from associations Vergesellschaftung, (Weber 1972). Associations, as formal organisations, are based on some form of rational purpose. By contrast, communities are primarily based on affectionate or traditional relationships as, for example, kinships, neighbourhood networks or religious sects. Affections can, of course, develop as latent functions in formal organisations, but are usually not the official purpose of formal organisations (Merton 1966). 3 Organisations define in their statutes various decision-making procedures and respective criteria. The staffing of boards, for example, is regulated according to which positions have the right to influence decision-making procedures. Examples are membership-based organisations with “democratic” appointment procedures or top-down appointments by individual owners. In addition, statutes sometimes specify certain procedures, such as veto rights or two-thirds majority demands for certain decisions. Furthermore, organisations define the decision-making procedures and criteria concerning how to staff their operations. The procedures define aspects such as who is allowed to influence staffing decisions and according to which criteria. Staffing criteria are differentiated for various functional positions in the organisation, for example, with respect to professional certificates or applicants’ personal affiliations. 4 Cooperatives, for example, pursue the welfare of their client-members by means of self-help. Philanthropic or charitable associations, on the other hand, dedicate themselves to altruistic welfare production for their clients. The latter can also be said about idealistic or denominational associations, but the activities of these associations often go beyond those of a single welfare association and, furthermore, such associations usually pursue general values in addition to those connected to welfare production. 5 This idea stems from the following studies. The approval of the labour movement for the accident insurance under Bismarck was gained through accepting that personnel recruitment was partly controlled by the social democrats (Rimlinger 1971). Other examples are political parties’ side organisations, such as party foundations, which are often used as an intra-organisational reward system (Panebianco 1988). 6 Kuhnle and Selle’s explanation is applicable up to the 1980s. Since then, they no longer see a partnership between state and associations. According to Kuhnle and Selle, the social and economical preconditions for statutory welfare have changed and resulted in a differentiated view of states on associations with increasing pressures to privatise and decentralise welfare provision. 7 Salamon and Anheier use the term ”social forces,” but they do not further elaborate upon it. However, they use the terms elites, classes and class elements in a similar way, thus indicating that “social forces” is a term that includes elites, classes and class elements. 238

8 Salamon and Anheier use the term ”landed interests” with which they predominantly seem to refer to aristocratic land owners. 9 Even in James’s example of denominational schools in the Netherlands, I am not sure if the religious groups were in competition with each other, that is, if they were really afraid of losing supporters to the other denomination. For the Roman Catholic Church, the competition with the atheistic labour movement may have been more crucial (see van Kersbergen 1995). 10 The discussion of separated statutory and associational welfare interests is flawed, because states cannot be regarded as entirely autonomous welfare agencies from associations. On the one hand, states influence associations through statutory decisions about transfers and regulations, but, on the other hand, states are also influenced by associations, which can act, e.g., as pressure groups to influence statutory decisions about , among other things, their role as welfare providers. 11The categories organisational capacities and socio-political complementarity refer to the common factors usually focused on in mainstream welfare research: first, organisational capacities, available for welfare agencies, second, functional welfare needs of welfare agencies’ members and, third, socio-politically mobilised interests of their social bases. Welfare agencies with centralised decision areas and well- developed standing administrations are regarded by some researchers to have higher capacities to allocate resources and to stipulate prescriptions, leading to greater welfare efforts by those agencies (e.g., Skocpol and Amenta 1986). Others regard broader socio-economic developments to have resulted in functional welfare needs, • which, in turn, influenced in which areas and for which social segments welfare became provided for (e.g.,Wilensky and Lebeaux 1958). Some scholars have instead highlighted the relevance of socio-political mobilisation for welfare development. Not functional welfare needs, but socio-political pressure from states’ support basis, i.e. pressure from mobilised and organised member groups, shaped welfare developments according to their welfare interests and ‘needs’ (e.g., Korpi 1981) The relevance of each single factor is debated among scholars, but one can instead consider in a first step that states and associations can be regarded as being different concerning all of these factors. It will be argued, that an association and a state differ concerning both their organisational capacities, which enable them to provide welfare differently, and their social bases, which pressure for different welfare needs and demands, leading to different welfare efforts. The categories here chosen, organisational capacities and socio-political complementarity refer to these mainstream factors, whereby socio-political complementarity is supposed to be the combined factor of functional welfare needs and socio-political interests: welfare needs, which are pressured for by mobilised members. 12The considered Western European countries comprise ‘EU 15’:Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxumbourg Netherlands, Portugal, Spain, Sweden, United Kingdom. Total tax revenues including contributions to social funds. Public social expenditure includes social spending by governments, publicly controlled social funds or social funds with compulsory contributions. 13Eventually, publicly organised welfare schemes do also attract volunteers. In Great Britain, e.g., ’volunteer centers’ coordinate voluntary work for municipal welfare (Lorentzen 1994). In other cases in Great Britain, however, volunteers even

239 shifted their activity towards other, non-public, controlled welfare activities, when welfare schemes became publicly organised (Finlayson 1994). 14 The level of donation is higher in countries where tax exemptions are allowed for: Great Britain has a high degree of donation/ generous tax reductions, while in the Scandinavian countries donation is low/ minimal tax exemptions, or, in Sweden, none at all. 15 Ideally, data about associational welfare before any kind of public intervention would have provided the best information about independent associational efforts, but, regrettably, such data is not available. Instead, data of associational welfare activities is taken at stages, when public intervention was on a limited scale: either through municipalities only without central state involvement, or limited in resources or time span. 16Again, the data available about associational housing is not from a time where public intervention had been completely absent. Housing associations had received minor public assistance in form of subsidies, tax reductions and easier accessibility to loans and land. In addition, major financial interventions from central states had taken place after the First World War, especially in Germany and the Netherlands (ILO 1930, Strömberg and Pooley 1992, Blumenroth 1975, Power 1993). 17 Overviews of housing policies in European countries give ILO 1930, Strömberg and Pooley 1992, Blumenroth 1975, Power for the time before the Second World War and Boelhouwer and Heijden 1992, Ball et al. 1988, Balchin 1996a, Danermark and Elander 1994a, Papa 1992, Oxley 1984 for the time after the Second World War; see also appendix A table A. 18Associations may, in addition, be attractive to some client categories, e.g., drug addicts or illegal immigrants, because associations were considered to be to a lesser degree connected to other public agencies, e.g., the police or immigration offices (Kramer 1989). 19 Cooperatives, for example, are also legally regulated with respect to internal decision-making processes (e.g., voting procedures) and goal definitions (e.g., interests of their members) (Fischer 1995). 20 Legal conditions for public acknowledgement differ widely between countries (Hansmann 1987). An example of legal differences is the classification of private universities and schools as charities in the USA and Great Britain. This classification entitles them to tax-deductible donations, which are an important financial resource for charities (Useem 1987, Ylvisaker 1987). The legal construction of schools and universities as charities regulates the success of these private organisations. Another example can be seen in the special recognition of cooperatives in Europe. Cooperatives with special statutes are acknowledged in many European countries as “socially oriented,” but not to the same degree as in the USA. Part of the definition is only, that associations, as formal organisations, underlie some form of legal regulation. How legal regulations are specified in a given country and what kind of criteria are formulated as conditions, are open research questions (in my opinion, even the acceptance of business organisations would provide interesting information about the political interests in welfare provision). Public acknowledgement is a crucial factor for welfare associations and entails being paid for the deliverance of services, to gain assisting grants for their activities

240

in general or to be exempted from various taxes (corporate taxes, real estate taxes, et cetera). Furthermore, associations’ donors become entitled to tax reductions of their income taxes. In Europe, public subsidies have often the form of direct payments for services or other forms of welfare provision. Besides these direct public subsidies, some countries, especially the USA and Great Britain, entitle welfare associations to tax-deductible forms of donations on a high level, which are very limited in Scandinavia (Lorentzen 1994, Gjems-Onstad 1995). 21Further conditions are, according to Lundqvist, that the implementing agency has to understand and has to be capable to act according to the intentions of the decision taker (Lundqvist 1987, quoted by Sannerstedt 2001). 22 Another reason for variations of welfare provision by associations is assumed to be the varying ability of associations to access resources and to organise welfare. 23 Another reason for statutory generous welfare efforts and associations’ more delimited ones could also be that a state is more capable of accessing financial means than associations. However, single associations not only had narrower welfare efforts than states, but also differed from each other in their welfare efforts, which is a further indication for a deliberate specification by associations. The wider welfare interests of states can also be seen in public welfare schemes existing parallel to associational welfare schemes in the same welfare area. Public administrated welfare schemes tended either to be less exclusive than associational schemes or to be ‘residual’ by providing welfare for those who were not able to cover their welfare concerns on the market or by associations. Regarding both • attempts, a recurrent pattern is, e.g., that client groups of poorer strata are concentrated in public schemes, e.g. municipal schools and day-care in Britain, public housing in Sweden and Norway. 24 This does not necessarily imply a statutory pursuance of a universal, citizenship-oriented welfare production, as the recurrent allowance of selective associational provision shows. States, for example, maintained their reliance on associations or left associations’ original operation intact. Social insurance providers in Continental Europe, for example, are subsidised and semi-publicly organised, but their original differentiation after occupational status remained intact. Other examples are associational operated schools and day-care facilities in Great Britain, which are allowed to demand tuition fees and to define rules for client selection. In other words, what is regarded by a state as ‘too narrowly particularistic’ associational welfare is relative and depends on, depending on the influence of particular group interests (social groups may be of such importance that their particularistic welfare concerns would be likewise ‘national' welfare concerns; see also next section over political parties). The point stressed here is rather that the statutory regulation of welfare associations repeatedly implied interference with associations’ originally more particularistic welfare provision in order to make welfare provision more nationally comprehensive, balanced and encompassing in a standardised way. 25 States and associations are not regarded to be influenced in the same way by social factors, which are usually regarded by welfare research as influencing (statutory) welfare efforts: neither are functional welfare needs evenly spread across the national population, nor are welfare interests socially pressed for evenly by all segments. Socio-economic developments, for example, affect to varying degrees

241 different segments of the national citizenry, which will therefore develop different functionally related welfare needs, according the socio-economic conditions for single social segments. Not all social segments e.g., were participating in the processes as of urbanisation or the femalisation-of-the-labour-force, leading to different welfare needs of, e.g., urban employees and rural populations or employed women and housewives. Furthermore, the capacities to mobilise resources for their interests vary between different segments of the national citizenry, which will therefore, to varying degrees, be able to press for or to organise their welfare interests. Well-trained workers and laymen, e.g., were well organised in movements and religious organisations in Scandinavia, which led to their ability to pressure and to organise their specific welfare interests to a stronger degree than, eventually, renting tenants or employed women. The social ‘unevenness’ of functional welfare needs and socio-political pressure is, in turn, assumed to lead to different exposures of states and associations to both functional welfare needs and socio-political mobilised welfare interests, due to their different membership and support bases. 26 Associations also employ non-public employees, who usually have a lower degree of job security and receive lower occupational rewards than public employees, who could enable associations to produce welfare at lower costs than public agencies could. However, only a few case studies have been made, showing varying results concerning associations' treatment of its personnel (Knapp 1989). Associations that are statutory funded are occasionally obliged to give quasi-public status for associational personnel (e.g. in Great Britain; Knapp 1989). Furthermore, the employment of non-public personnel is a 'capability' that is not only available for associations, but also for other private organisations, e.g., business firms. 27 The ideology of liberal parties and conservative parties leads to a preference for associational organisation either to avoid a “strong state” or to support “pre- modern,” corporatist organisation by status differentiating and Church-related associations. One could interpret this notion as indicating that these researchers assume that liberal parties are interested in low government welfare spending and in efficient or quasi-market provision of welfare by associations (Salamon and Anheier 1998). 28 In this section I will focus on interests in owning an associational welfare organisation in order to control the operation of welfare. General interests mentioned in the introductory section, such as owning a welfare association in order to enhance social prestige or to gain opportunities for patronage or elite training, are, in this introduction to the welfare field of day-care, not further elaborated upon, because interests in social prestige and organisational infrastructure are of more general character, i.e. not necessarily related to the welfare area of day-care. 29 The development in Sweden and Norway was no exception, where “barnasylums/barneasyl” were founded in Oslo, Stockholm, Gothenburg and Bergen around the beginning of the 20th century (Svenska fattigvårds-och barnavårds kalender, NOU 1972:39). De Nödlidendes Venner, for example, founded the first asylum for children in Trondheim. Their intentions were “…at åbne et Asyl for Småbarn hvis fattige Foreldre skulde ved Arbeid udenfor hjemmet vinde Brödet for seg og sine …for at forebygge den Skade Börnene lide ved at overlades til seg selv eller udsette dem for den Fordervelse om Örkeslöshet, Tiggeri og Omgang med

242

Vanartade uunngåelig medföre.” “Fra disse beretninger får en inntryck av at asylbevegelsen hjemme och ute ble båret fram ved at idealistiske, kunnskapsrike og velstående mennesker så sitt ansvar for medmennesker og foretok seg noe for å avhjelpe nöd blant barn…” (citet of Tora Grunde,in: NOU 1972:39). 30 Occasionally, day-care facilities that had been established by associations were actively assigned to the municipality (Hatje 1999; Helgesson 1995, see also below). 31 In order to do so, I have used the indices of the Swedish law collection SFS and the Norwegian parliamentary debates SF. Furthermore, administrative prescriptions have been used from the Swedish National Board of Health and Welfare (Socialstyrelsens) (Råd och anvisningar, Socialstyrelsen anser, et cetera) and the Norwegian government. 32 These negative effects could be due to associations’ inability to raise tax resources or limited ability to access further donations and voluntary welfare work in order to finance costly operational standards (see introduction). 33 Increased parent fees would also be an option for associations to increase their funding. Both countries, from the beginning of their day-care policies, prescribed “reasonable” fees as a precondition for statutory funding. Furthermore, they both introduced maximum limits for day-care fees, which limits associations’ ability to raise fees (2003 in Norway, 2001 in Sweden). According to more recent studies in Sweden in the 2000s, no large differences between municipal and associational fees exist. Associations predominantly use the same payment schedules as municipalities and even rely on municipal • administration of fee collection (Skolverket 2003:231, 2005:261). On the other side, in Norway, associational parent fees have been higher than average municipal fees. They are at the upper limit of allowed parent fees (Rauan 2005). 34 The National Board of Health and Welfare (Socialstyrelsen) was responsible for examining and deciding upon applications by associations (in the 1940s, Labour Market Councils also had to approve the need for day-care facilities in a municipality; Råd och Anvisningar 1944). Furthermore, the National Board of Health and Welfare issued specifications of legal regulations, which partly had a binding character like administrative laws, but were partly non-binding recommendations. Examples for binding requirements are educated heads of day-care facilities, ratios of the number of children per pre-school teacher, children group sizes, standardised hygiene routines, regular health controls, standards of room sizes, equipments, and so on (Socialstyrelsen Råd och anvisningar 1944, 1949, 1963). Exemptions from required standards were often allowed for. 35 Kindergartens (lekskolor) usually had opening hours of up to 3 hours per day (Råd och Anvisningar nr. 163, 1963). 36 I have calculated the amount of subvention paid per place by setting salary subsidies in relation to officially recommended maximum sizes of children group per teacher in Kindergartens and other day-care forms for 1959 and 1962. Since 1963, subsidies were officially specified per place. 37 In Norway, by comparison, subsidies for short-time day-care remained at higher levels until the beginning of the 1980s, when they were likewise set to 10% (see appendix A table A).

243

38 It was not expected that exisitng kindergartens would have difficulties with adjusting to a lowered subsidy levels, because former levels of subvention had been relatively generous (Statutskottets utlåtande nr. 76, 1966). Objections were raised by the Swedish Pre-School Teachers Association (Sveriges Förskollärares Riksförbund). 39Indirectly, the decreased support for short-time day-care also implied a shift to day-care with more demanding operational requirements, because kindergartens had to comply with far less demanding regulations than half- and full-time day-care facilities. The children per teacher ratio, for example, was 1 to 20 in Kindergartens, but 1 to 5 in half-and fulltime day-care in 1962. 40 At the same time, the statutory subsidy system was changed from ear-marked benefits for childcare to block grants for all types of municipal expenses. 41 In combination with the introduction of unspecified block grants for municipalities as a replacement for former earmarked grants from the central state. 42These recommendations could, for example, not prevent considerable increases of parents’ fees during the 1990s (Blom-Hansen 1998). 43 Extra subsidies for transforming day-care provision were granted in connection with the lowering of the school age from 7 to 6. 44 The ability of public providers to provide longer opening hours and care for younger children even during the 1980s, when statutory funding was not generous for such day-care, may be due to their ability to add their own resources. 45 All day-care providers, municipal or associational, were allowed to stipulate a “basically Christian orientation” – kristen formålsparagraf . Such a general Christian orientation was, however, limited to a moderate interpretation and still required cooperation with the children’s parents (Day Care Acts 1995, 2005). 46 Other requirements addressed the physical environment, such as area per child, heating, playground equipment, and so on. 47Exemptions from the employment of trained personnel are possible in both countries. The argument against strict regulation was, and is, that the availability of trained pre-school teachers often falls short of the need for personnel to fill existing positions (Ot.prp.nr. 75, 1980-81, St.meld.nr.27 1999-2000, Ds S 1965:3, Socialstyrelsen anser 1978). Both countries are engaged in financing pre-school training and education for day-care providers at colleges and universities in order to enhance the availability of pedagogic personnel. 48 Parent fees should provide for sibling moderations, but not necessarily for income moderation as in Sweden. Furthermore, providers could be exempted from parent fee limitations in cases where their operations and the quality standards of their operations would be endangered; however, exceeding the maximum fees had to be approved by parents (Forskrift om foreldrebetaling i barnehager 2003, Day Care Act 2005). 49 It might be assumed that funding for short-time day-care would favour associational provision, due to associations’ predominant occupation with short-time day-care during the first decades of state funding. However, the development of associational involvement seems not to support this assumption entirely. However, proportionally decreased subsidies for short-time day-care in Sweden in 1963 were followed by a decrease in activity by associations, which had been the foremost providers of short-time day-care until the 1960s. Thus, diminished funding

244

for short-time day-care could have been the reason for further associational decline. In contrast, proportionally decreased funding for short-time day-care seems not to have discouraged associational engagement in Norway. Despite associations’ predominant activity in short-time day-care during the first decades and decreased funding for short-time day-care between 1982 and 1997, associations maintained their high provision share in Norway. Likewise, professional standards for day-care personnel seem not to have inhibited the growth of associational day-care provision in Sweden. 50As providers of day-care, associations are supposed to be able to influence several operational aspects, such as selection of personnel or clients, length of opening hours, specification of curricula. Associations’ interest in operating day- care facilities themselves may depend on how eager their founders are to control such aspects of operation. 51 The category of unspecified owners consisted at least partly of single, pedagogically trained women. They had an interest in providing pedagogical day- care, but their economic situation was precarious. Pedagogues complained regularly about the difficult financial situation of day-care institutions and seemed, therefore, to have preferred public employment (Hatje 1999). 52 Associations are officially involved in the legislative procedure. In what is called the “Remiss” process proposed bills are circulated for comment to all groups concerned, and eventually modified before the final version is enacted. Associations therefore have the opportunity to publicly and officially comment on legislation that • concerns their activities. 53 Two illustrations can be given: Helgesson investigated the same question and found no hint of protest in City of Gävle; not even when former associational day- care was overtaken by the municipality (Helgesson 1995). An investigation into the municipal takeover of the associational facilities of the most famous pioneers of Swedish day-care shows a similar acceptance (Ellen and Maria Moberg founded several day-care facilities in Norrköping and a pedagogical institute for day-care personnel, Fröbelinstitutet). The board of the Moberg heritage even applied for municipal takeover (the Mobergs themselves were worried about social-democratic influences on their facilities, but they, in principle, advocated municipal engagement; Hatje 1999). 54Norwegian statistics are available for several categories of associational provision :day-care facilities of employers, housing cooperatives, parent cooperatives, individuals, foundations, private organisations, parishes, joint-stock companies, foundations and others (see table categories of associational providers, appendix A table A). 55 ”Tiden är inne til å satse på förskolebarna I våre menigheter. […] Grip sjansen nå! Om noen år vil den offentlige förskolen blir en virkelighet også i vårt land og da har vi som menighet redusert muligheter for å påverke disse […] Tal saken i menigheten: Mulighetene er store!”(Norwegian Sunday School Union, quoted by Lea 1982). ”[…] Men dette betyr også at det er nå vi har vår gyldne anledning til å få eist kristne barnehager, om et par år kan det väre for sent” (Norwegian Luthersk Misjonssamband, in”Vi bygger barnehage,” quoted by Lea 1982).

245

The day-care consultant of the Norwegian Christian umbrella organisation, Institutt for Kristen Oppseding, underlined in the same way the necessity to rapidly establish associational Christian day-care, otherwise municipals would entirely dominate day-care (press interview 1979, quoted by Lea 1982). 56 Parents with children in other types of facilities named close-by location or good overall impression as the reasons for their choice. The study indicates further that parents in parent cooperatives are distinct in terms of their socio-economic position. They are more often self-employed and higher civil servants, while a working-class background is less usual (Sundell and Ståhle 1996). This could be interpreted as a wish to obtain care for their children that is distinct from publicly provided care. 57 A strict comparison over time is difficult, due to changing categories in public statistics. Still, the interpretation that private organisations, parishes and denominational associations’ shares decreased in comparison with parental and individually owned associations is considered to be reasonable, because these former associations were still predominant in 1990, when parental and individually owned associations were already part of public statistics. 58 Long-time day care for younger children is more expensive than other forms of day-care. Employers and parent cooperatives may have been better able to provide expensive day-care, as they had access to more resources: employers via their companies and parent cooperatives via access to voluntary work by parents. In Sweden, no differences exist between municipal and non-public providers concerning the length of children’s attendance (Skolverket 2005:261). 59 In neither country did tighter operational regulations seem to hinder associations from engaging in day-care provision during the second phase. The introduction of parent fee limits, however, may decrease the social segregation between parents with children in municipal day-care and those with children in non- municipal day-care. 60Information about political parties’ positions is gathered from official investigations, government drafts, drafts of parliamentary committees and parliamentary debates. 61 Only the women’s group of the Conservative Party emphasised that, in principle, children should be taken care of at home. Nevertheless, they acknowledged the “great need” for childcare even outside the home (Prop. 1943:339). 62 “Att ge de enskilda förskolorna samma statsbidrag som de kommunala får, skulle innebära en pedagogisk injection när det gäller hela förskolevrksamheten.” “Vidare anser vi att det tillhör föräldrarrätten att kunna välja en alternativ förskola, för att t.ex. tillgodose speciella behov hos barnet eller för att få en speciell pedagogisk eller ideell inriktning.” “Valfrihet måste medges” (Ann-Cathrine Haglund; Conservatives, Prot. 1982/83:126). “ Rune Gustavsson signalerade en frihetskamp mot byråkratin och stora svängrum för den enskilde.” “Det kommer att ge en bättre regional spridning…och det tar också i anspråk det kapital av ideell kraft som bör utnyttjas, och det kommer att ge utrymme för stor

246

variation och uppfinningsrikedom” (Ulla Tillander, Centre Party, Prot. 1982/83:126). “Det är många tusen timmer av frivilligt arbete som lags ned på administration, på städning osv., och där är det inte förbjudet för föräldrarna att göra en ínsats för det egna barnet” (Ulla Tillander, Centre Party, Prot. 1982/83:126). […] “man kan få den billigare utförd, och man kan dessutom dra nytta av erfarenheter från alternativ pedagogik, som snappas upp fortare enskild verksamhet.” (Ingemar Eliasson, Liberal Party; Prot.1982/83:126). 63 “Syftet med moderaternas privatiseringskampanj är, misstänker jag, att man i ett kärvt ekonomisk läge vill ta pengar från den kommunala barnomsorgen och slussa över dem till privata daghem för högavlönade.[…]Vi är öppna för att pröva nya former för att förbättra föräldrasamverkan, stödja föräldrakooperativ och olika sammanslutningar för att göra insatser inom barnomsorgen. Den verksamheten bör uppmuntras, om den nämligen uppfyller kraven på rättvisa och god kvalitet.” (Sten Andersson, Social Democratic Party, welfare minister, Prot. 1982/83:126). 64 Social democratic governments excluded for-profit providers from public subsidies when they returned to power in 1994, but that lasted only for two years. 65 St.meld.nr.89 1961-62; Innst.S.nr.11 1962-63, BudsjettInnst.S.nr.11 1969-70. 66 See appendix A table A for partisan composition of governments and parliament majorities. 67 From this perspective, municipalities should regard private day-care facilities mainly as a supplement to a municipally owned and organised day-care system • (“Private barnehager bör kommunene först och fremst se på som et supplement til et kommunalt eiet og organisert barnehagesystem;” St.meld.nr.8 1987-86). 68 J. Östtveit, KRF: “Eg trur likevel private organisasjonar gjennomgåande gjer eit slikt arbeid lettare, smidigare og meir effektivt. Og er ikkje nettopp den private innsatsen med og bremsar tendensen til å skuve frå seg medmenneskeleg sosialt ansvar.” Aarvik, KRF: “Jeg tror det er riktig å sates på samarbeid mellom de private organisajoner og kommuner og stat - et samarbeid som jo har gitt så rike resultater.” (SF 28.nov. 1969). 69 “Jeg vil i förste omgang bare spore statsråden om ikke utviklingen bekymrer. [...] En ting er rotet omkring formålsparagrafen, som kunne ha vaert unngått. En annen er at mange kommuner setter bort til private kristne institusjoner den barnehageutbyggingen som ved stortingsvedtak er kommunenes klare ansvar. Dette vil i mange tillfeller si at det er å sette bort utbyggingen til organisasjoner som egentligen er imot barnehager, og som i förste rekke er opptatt av formålsparagrfer og prioriterer deltidsbarnehager framfor andre utbyggningstiltak.” (Haldis Havrøy, Labour Party; SF 28. feb 1979). 70 “Kari Lie viser til den frie yrkesutdannelsen I vårt land, og hevder at det ville bli et rent narrespill hvis barnehageeiere skal få lov til å stille krav om personlig livssyn ved ansettelse. For egen regning vil jeg si at skulle dette bli tilfellet, ville Norges storting også gjennomföre et rent yrkesforbud. Det er menneskekrenkende å forlange beskjennelsesplikt for enkelte yrkesutövere.” (Haldis Havrøy, Labour Party; SF 28. feb 1979). 71 “Det henstilles til Regjeringen å sörge for at forskriftene til barnehageloven endres snarest mulig slik at en får klargjort foholet mellom lovens formålsparagraf

247 og eventuelle livssynsvedtak. Det må klart framgå at direkte forkynning og undervisning ikke harmonerer med lovens intensjoner.” (Haldis Havrøy, Labour Party; SF 28. feb 1979). 72 “Jeg vil i förste omgang bare spore statsråden om ikke utviklingen bekymrer. En stor og viktig sosialpolitisk reform er nå i ferd med å klattes bort.” (Haldis Havrøy, Labour Party; SF 28. feb 1979). 73 “Når det gjelder de private barnehagene, er det klart slått fast i Stortinget at de skal ha frihet til å fastlegge retningen for sin barnehage sjöl, så sant det er i samsvar med loven. Kristelig Folkeparti er for sin del fullt beredt til å ta opp en kamp, hvis det er nödvendig, for å sikre de private barnehagene same rett som friskolene har på skolens område.” “Jeg önsker å understreke at personale som skal ivareta det pedagogiske ansvar for barnehagene, er lojale of utförer sin funksjon I tråd med barnehagens förmålsparagraf.” (Kjell Magne Bondevik; Eli Kristiansen; SF 28. Feb 1979). 74 The ministry “… vil söke å styrke utbyggningen av halv-og heldagsbarnehager […]” and “[…] barnehagerplasser for barn i aldren § til 1§ år […],” because “[…] foreldre önsker […] dersom den skal kunne tilby tilsyn med barna mens begge foreldrene er i arbeid utenfor hjemmet.” (St.meld.nr.93 1980-81). 75 “Mindretallet vil I denne sammenheng understrecke at eierne av private barnehagr må ha frihet til å tilsette personale som deler det livssyn insitutsjonen harlagt til grunn for virksomheten. Kravet om at den ansatte skal vaere faglig kvalifisert, må vaere det eneste som kan overpröves av barnehagenemnda eller andre” (Innst.S.nr. 367 1980-81). “Mindretallet har vidare market seg at meldningne gir inntryck av at allmenne barnehager med Kristen formålsparagraf vil få så sterke restriksjoner for praktiseringen av denne, at den pedagogiske og metodiske frihet nesten går tapt. De såkalte ikke-allmenne barnehager gis derimot en frihet som er betinget av en så sterk ekonomisk egeninnsats at de faerreste barnehager kan makte å drive på slike vilkår.” (Innst.S.nr. 367 1980-81). 76 “Stortinget har vedtatt en endring av §55A i Lov om arbeidervern som får virkninger for ansettelser i barnehage. Etter dette er det tillatt å spore sökerne om deres livssyn dersom det fölger av stillingens karakter eller virksomheten har er bestemt formal av f.eks. religiös art. […] Departementet vil understreke at eiere som har fastsatt en vid fortolkning av lovens livssynsformål, ikke skal pålegges noen form for ökonomiske restriksjoner.” (Ot.prp.nr. 75 1981-82). 77 In addition, they introduced Christian principles for all day-care providers, including municipalities. The difference between providers promoting general Christian principles and those whose specific rights were upheld is that the former providers must still operate in accordance with parents’ preferences and they cannot ask job applicants about their convictions (Ot.prp.nr. 75 1981-82). 78 “Det kan imidlertid vare behov for en store grad av frihet for enkelte barnehager som er opprettet med et mer direkte religiöst siktemål. Mange av menigheter innen Den norske kirke og private organisasjoner og menigheter driver barnehagearbeid som i tilltak til et sosialt formal, har som henskíkt å gi et tilbud med et klart kristent innhold.” “Dersom foreldrene söker plass i barnehager som har en slik vedteksbestemmelse, vil den begrensning av praktiseringen av livssynsformålet

248

som ligger i“samarbeid med barnas hjem,” ikke lenger foreligge.” (Ot.prp.nr. 75 1981-82). 79 “[…] för att störst mulig antall barn skal få nyte godt av det pedagogiske og sosiale tilbud barnehagene er […]” (SF 1982, 21.april, Fredriksen, KRF). 80 According to Lea, “religious-conservative” associations were predominantly organised by the Housewives’ Federation, Vesterlandske Indremisjonsforbund, Det Norske Misjonsselskap, Det norske Lutherske Indremisjonsselskap, Den norkse Santalmisjon and parishes of the Norwegian State Church (Lea 1982). 81 It could be assumed that associations have day-care provision in areas where they have strong membership bases offering ready access to affirmative clientele. However, the prevalence of day-care facilities is a flawed indicator of clients’ wishes for specific day-care in times of scarce day-care provision. It will be used in the absence of more direct data about parents’ wishes for such forms of day-care. 82 Instead of the regional dispersion of constituencies, the use of individual-level data would have been more precise, but sufficient data are not available. The risk of an ecological fallacy is apparent, inferring wrongly from geographical support to individual support (Freedman 2004). Furthermore, the risk of drawing wrong conclusions is heightened by the fact that the support for both the associations and the political party represents only a small fraction of the regional populations. Efforts have been made to limit the problem of confounding individuals’ wishes for day-care with regional dispersions via inclusion of normative orientations from the organisations involved. • 83The “Southwest” refers to six counties (fylke), including Aust-Agder, Vest- Agder, Rogaland, Hordaland, Sogn og Fjordane, Möre og Romsdal, which comprise about one-third of the Norwegian population. These counties are situated at the southern and western coastline of Norway, beginning west at Oslo and stretching towards the mid-line of the country. 84 Historically, the Christian People’s Party was a splinter group of the Liberal Party, which was based on a socially and culturally wide-ranging coalition (Rokkan et al. 2001). The break-off of the Christian People’s Party occurred due to conflicts around the liberal cultural orientation of the Liberal Party, for example, over questions of religious education in schools and liberal freedoms in the arts (Bondevik 1975). 85 ”Kristeleg Folkeparti vil verna um dei kristelege og nasjonale verdiar i kyrkja, skule arbeidsliv og i vårt kulturliv i det heile” (Prinsipprogram 1933). Most Norwegian parties adopt two kinds of programs: basic programs (prinsipprogram) and action programs (handlingsprogram). The former are general statements of ideology and long-term objectives. They are adopted by national party congresses (Strøm and Leipart 1989). 86 The Norwegian State Church also organised denominational day-care associations, but due to high membership rates of Norwegian citizens in the State Church, membership data provide only minor information about strongholds of social support. The high degree of overall religious activity in the Southwest, however, indicates that State Church organisations can also rely on considerable support from the population in the Southwest (see text above).

249

87 The Vestandlanske Indremisjonsforbund, for example, was founded by lay activists in 1898. Det norske lutherske Indremisjonsselskap was likewise founded by a lay movement in 1868. 88 The Vestandlanske Indremisjonsforbund, for example, has seen its main task as preaching and doing missionary work. ”I utgangspunktet så Indremisjonsforbundet det som sin hovedoppgave å sende ut forkynnere med Guds ord til folket.” Normisjon, which is a fusion of Indremisjon and Santalmisjon, acts under the guiding principle of “reaching the unbelieving and turning them into disciples” (”FORMÅL: Normisjons visjon er å nå de unådde og gjøre dem til disipler”). 89 “Vår visjon: Vi vil gjøre mennesker til Jesu disipler ved å gå med evangeliet til våre nærmiljø og til folkeslagene. Vi prioriterer å bygge lokale åndelige fellesskap som basis for liv og tjeneste. Vi prioriterer medmennesker i vårt land som ikke kjenner evangeliet. Vi prioriterer å gå til de unådde folkeslagene som er uten det kristne vitnesbyrd på sitt språk. Vi prioriterer diakoni og omsorg i møte med mennesker. [...] Medarbeidere i Normisjon å være forpliktet på den evangelisk-lutherske bekjennelse og de verdier som organisasjonen har som grunnlag for sitt arbeid.”(Normisjon). 90 ”Ansatte har også i høy grad kombinert forkynnerarbeidet med diakonalt arbeid.” (Indremisjonsforbund). 91 This view of day-care was repeated in detail during the second half of the 1980s, when economic difficulties of part-time day-care were criticised. 92 In 1998, they renamed themselves the Norwegian Women and Family Organisation (Norges Kvinne- och familieforbund). Furthermore, they changed their basic program in 2000 (after having formulated a new program in 1997) without repeating their Christian orientation (Prinsipprogram 2001, Fjerstad 2001). 93 The so called “kontantstötte” offers all parents with children aged 1-3 years who do not attend state-sponsored day-care the same amount of money as the state subsidy per place. After years of dispute, a right-wing coalition introduced the scheme, against the votes of the opposition, in 1998. 94 The Housewives Federation supported the home-care allowance scheme, but opposed the name of the program. They wanted to call the payment a wage instead of a benefit (Housewives’ landsmötet 1993). 95 Vote shares and provision shares had to be categorised because of the scarcity of day-care facilities. A problem with the data for 1980 is that over half of all municipalities had no more than two facilities in total. Furthermore, over half of all municipalities had no religious-conservative day-care. 96 I have not further investigated, if right-wing parties and associations shared or not congruent constituencies in Sweden until the 1990s. 97 Tenant has a twofold meaning: it can mean occupant, but also more specifically occupant of a rented dwelling. I will use the term in the sense of occupant. If I refer to tenants in rented dwellings, I will use the term renting tenant. 98 In the wake of industrialisation and urbanisation, preconditions for housing construction changed in several ways, e.g., the development of land markets for urban building sites, technical development of cost-intensive multi-dwelling houses

250

and financiation through mortages, making interest payments an important part of housing costs and connecting housing costs to capital market developments. Furthermore, most housing was no longer self-built but instead constructed by professional investors. Labour had to be contracted for construction work, and wages became added to development costs. In addition, investors build houses for assumed market demands and profitability. The increased market influences on housing resulted in volatile building costs and volatile building activity (ILO 1930, Strömberg and Pooley 1992). Market influences not only altered preconditions for housing development, but also ownership and tenure of housing. In urban areas, where multi-dwelling housing was predominant, rented housing by private landlords became the usual housing form (ILO 1930, Strömberg and Pooley 1992). In rural areas small-scale, home- owner housing prevailed during the first and, to a lesser degree, second half of the century (ILO 1924, 1930, Population and Housing Census 1933, Population and Housing Census 1945). 99 Public authorities also intervened to alter the ownership structure of housing. For this purpose, they provided subsidies for limited-profit organisations and tax redemptions for owner-occupiers. In addition, public authorities intervened through regulation of the housing market. Market organisation of housing was reduced in several ways: concerning land use through increased municipal land control, concerning sanitation problems and overcrowding, and through quality regulations and rent restrictions (ILO 1924, 1930, Strömberg and Pooley 1992). • Public involvement reached a first peak during and after the First World War, but gained real importance after the Second World War, when most European countries introduced large-scale housing construction programmes and subsidy schemes to alleviate shortages and housing standards (Boelhouwer and Heijden 1992; Balchin 1996a). The post-war impetus behind housing policy throughout Europe was reduced during the 1970s and 80s, and political intervention in the housing market shifted from investment support to policies of tax redemptions for owner-occupiers and rent and tenure regulations (Strömberg and Pooley 1992; Boelhouwer and Heijden 1992). 100 Here, the focus will be on ownership of houses through associations, combined with other activities such as development, administration, banking, construction, etc., but associations became involved in various aspects of housing, ranging from banking to construction to development and ownership. Examples of associational saving banks are British Housing Associations and German Bausparkassen, which both provided mutual banking for building purposes. Also, the Scandinavian owner-cooperatives act partly as savings banks. Associational construction firms included, for example, employee cooperatives, organised by construction worker unions. 101 ILO estimates that statutory subsidies and loans assisted about 80% of all new buildings in Norwegian towns (1914-28, five biggest towns). According to the ILO report, this was followed by an increase of associational initiatives in the five biggest towns from 1914-28. In the biggest towns, housing societies built 29%, public authorities 47% and private investors 24% of new buildings from 1914-28 (ILO 1924, 1930).

251

The first census of Norway that provides detailed information about associational housing dates from 1960: 11% of all dwellings had an associational tenure, while the remaining dwellings were privately organised (36% privately rented and 53% owner-occupied). 102 Comparing instead only larger municipalities (with at least 10 000 inhabitants), data suggest that associations’ shares of housing accommodation were 10-11% in 1920/1933 and 18% in 1945. 103 Statutory loans were issued for the first time in 1917 (ILO 1930, Strömberg 1992). 104Legal eligibility was regulated differently, but in general was rather unspecific about investor categories (SFS 1933:558, SFS 1941:636 and SFS 1942:569, to mention some central schemes). 105 Subsidies and loans included in this analysis aggregate central state and municipal aid. Municipal subvention went to 66% of all dwellings supported, while central state subvention went to 80% of all dwellings supported from 1931 to 1945 (Population and Housing Census 1945). 106Public utility companies were semi-public, including philanthropic associations and municipal authorities. Cooperative housing was entirely associational. 107 The age of their housing stock is only a rough indicator of the year the property was acquired. Associations and semi-public investors usually built the houses and dwellings they owned themselves (75%). Nevertheless, the remaining third of the dwellings they owned was bought. 108 Employer housing is not specified separately in the statistics for public aid, but is included in the category “private housing.” Since 1933 private employers were no longer listed as “public utilities” in the statistics. Employers accounted for 1% of all newly built dwellings from 1931-45. 109 “Husbanken skall ikke finansiere ukontrollert och spekulasjonsbetonet byggevirksomhet” (Innst.O.I 1946); ”Det framstår som angeläget att bostadsbyggandet i fortsättning till huvudsaklig del kommer till stånd i regi av icke- spekulativa företag” (SOU1945:63 I). 110 Rent regulations were applied for the first time in 1915 and 1917 in Sweden and Norway, but were not upheld over a longer time (ILO 1930). 111 Of the former Swedish housing policy, only collective rent bargaining is upheld. 112 Additionally, guaranteed interest rate levels were raised more quickly for home-owners than for other investor categories. 113 Percentage levels varied somewhat over the decades, but always included a “ranking” between public utility companies and cooperatives, on one side, and private investors on the other side (levels for municipalities remained 100%/30%, cooperatives varied between 99-8%/29-8% and for private investors 85-95%/15- 25%). During the 1940s, coverage ranged from 90 or 95% and loans from 20 or 25% of the development costs, with private investors receiving at the lower level. 114 Likewise, taxation levels for municipal and cooperative investors remained more favourable than for private landlords.

252

115Additionally, tax levels became equalised between all investor categories (Boverket 1994). 116 Before the Second World War, some municipalities intervened in the housing market. On the national level, policies were circumscribed to some years around the First World War. Occasional subsidies and rent regulations were abolished after some time (Bostadsstyrelsen 1968; ILO 1930). 117 In Sweden, direct state loans were only meant to cover a maximum of one- third of developing costs up to a certain share of those costs (second-priority loan). The Norwegian solution was aimed at providing the first- as well the second- priority loan (totalfinansering). 118 Private investors are included in the category “private” because they built mainly for owner-occupiers, i.e. they usually sold their buildings to owner- occupiers. Private landlord investors are not specified concerning loan issuing in official statistics (State Housing Bank information; St.meld. nr.63 1967-68). 119 Besides independent cooperatives, shareholder companies were also chosen as a form of collective ownership. Shareholder companies are still active in Norway and operate under comparable legal regulations as independent cooperatives. 120 Housing cooperatives act otherwise under the same regulations as business firms. 121 Prices did not rise immediately after the abolishment of price controls. On the contrary, they remained very low during the 1970s. This changed in the mid-80ies, when prices started to rise due to deregulation of the credit market in 1985 and due • to that banks accepted cooperative dwellings as collateral for long-term loans in the same way they accepted single-family dwellings (Turner 1998, Victorin 1987). 122 Before 1995, municipalities could acquire single dwellings within cooperatives, but had to request permission from the cooperatives. Because cooperatives were reluctant to permit such acquisitions, a right for municipalities and employers was introduced. 123 Another possible explanation is municipal influences on building. Municipalities decided about access to building sites and may thus have influenced the relative building activity of investor types. 124Until the 1980s, both left-wing and right-wing governments extended housing- related subsidies. In Norway, the right-wing coalitions from 1965-71, for example, raised guaranteed interest levels slightly and introduced a value-added tax on housing construction, but they also enhanced public loans and introduced a “building site subsidy” (Reiersen and Thue 1996, Annaniassen 1996a). In Sweden, the right-wing alliance from 1976-81 reduced subvention for older loans, but allowed more generous tax deductions for home-owners (Lundqvist 1992a). 125Since the 1980s, it is not only right-wing governments that have tried to limit housing-related expenditures. The Swedish social democrats, for example, decreased subventions for older loans and introduced several housing-related taxes during the 1980s. In cooperation with the Liberal Party in 1991, they sharply increased VAT taxes and property taxes, reduced interest subsidies for all housing investors and lowered deductibility of interest payments for home-owners. The coalition of bourgeois parties that followed maintained the higher level of taxation and further reduced subsidies (Papa 1992, Boverket 1994). In 1982, the Norwegian social democrats introduced reductions in public loan issuing and loan cost coverage, a

253 reduction of interest rate subsidies and the abolishment of the “building site subsidy” (Nagel 1992, Annaniassen1996b). Furthermore, the social democrats set the interest rates of public loans at market levels and thus abolished interest rate subsidies in the 1990s. These changes were supported by right-wing parties. These proposals were then re-affirmed by a right-wing coalition after the elections 1981 and 1996. 126 Preliminary regulations for cooperative tenantship were enforced by the Social Democratic Party in 1986 and 1999 (SFS 1986:1242; SFS 1999:459). 127 The main aim is to give tenants greater influence over the administration and operation of their dwellings and, further, to enhance social interaction between neighbours (SOU 2000:95, 13). 128 The Social Democratic Party wanted to “support building activity of low-cost rented housing and student housing.” (“Stötta byggandet av hyresbostäder och studentbostäder till rimliga priser;” SAP Party Program 2005). 129 The Conservatives had fought a decade-long battle for private landlords and free-market housing in Olso before the war, but altered their position after severe losses in municipal elections (Gulbrandsen 1980). 130 The governing Labour Party introduced a reduction of public loan issuing and cost coverage, which was supported by the opposing right-wing parties in 1982. The State Housing Bank would should provide loans in general only up to 60% of project costs, but should provide an additional loan of 15% for cooperatives under the condition that sale and price restrictions were guaranteed (St.meld.nr.12 1981- 82). In 1995, the governing Labour Party abolished interest rate subsidies for public loans, a retrenchment that again was supported by right-wing parties (Reiersen and Thue 1996). 131 From 1983, these parties formed a coalition. From 1981 to 2005, right-wing parties held a majority position in government. 132 Legal prescriptions about access to independent cooperatives were introduced unanimously in 1977 and abolished by right-wing parties in 1981 under the protest of the Labour Party (Gulbrandsen 1988; Annaniassen 1996c). 133 In 1995, the Labour Party had already stipulated amendments to the Cooperative Housing Acts, which allowed municipalities to purchase up to 10% of dwellings against the will of cooperatives. 134The Labour Party included public rented housing as a target in party programs since 1989: “…increased building activity for non-commercial rented housing and low-price owner-occupied housing for the young and less affluent.” – (“Öke byggningen av ikke-kommersielle utleieboliger og lavinnskuddsboliger for ungdom og ökonomsik vanskeligstilte”en av tre huvudpoäng, bredvid fler bostäder och ökad bostötte till barnefamiliene,” (Fjerstad 2001, Det norske arbeiderpartiet). 135However, one has to keep in mind that statutory housing policy is only one factor for the relative success of investor categories, besides household demands or input factor costs (see also section 3). 136 OBOS (Oslo Bygge - Och Sparelag) is the biggest Norwegian regional cooperative. In Norway, regional cooperatives played a more important role than in Sweden. NBBL was formed as an umbrella and service organisation in Norway and was not a direct developer of housing projects. 137 The national cooperative organisations were formed later from already operating cooperatives, such as umbrella organisations.

254

138 On the international level, Norwegian and Swedish cooperatives are members of ICA (International Cooperative Association). 139 Urbanisation and the accompanied housing situation was a main factor for the development of the cooperative movements. Swedish and the Norwegian cooperatives started their activity in the big cities. Stockholm and Oslo were the foundation places for HSB Stockholm 1923 and OBOS 1929, both foundations were directly connected to modern urban housing problems (see appendix A table A). 140 In both countries, several “closed” cooperatives had started to build houses, but stopped further activity once they had completed their houses. Those smaller projects had no further influence on the wider housing situation in the cities. Furthermore, they were usually founded by rather affluent groups. OBOS was first founded as a closed cooperative, but reorganised according the HSB model via a contract with the municipality of Oslo in 1935 (ILO 1964). 141 HSB statutes 1. paragraph (citet in Lundevall 1992: 54). “Vi bevakar riksyggensföreningars interessen, verkar for bostadsrätten som boendeform och för allas rätt till ett bra boende” (Riksbyggen). 142 “Bostad är en grundläggande rättighet. En god bostad är avgörande för möjligheten att leva ett gott och tyggt liv” (SAP party program 2005). “…att lånen och subventionerna komma en förbilligad och förbättrad bostadsförsörjning och icke spekulativa enskilda interessen tillgodo” s. 529; “… att undanröja den spekulativa karaktär, som[…]präglande bostadsbyggandet och bostadsförvaltningen” s. 534; SOU 1945:63 I. • 143 “Boligbyggingen må skje etter sosiale prinsipper ved samvirkende boligselskaper og med offentlig støtte. Staten og kredittinstitusjonene går sammen for å muliggjøre en boligkreditt som alle lag av folket kan ha nytte av. Ved garantier eller felles innkjøp og produksjon sikrer det offentlige en rasjonell fremstilling og omsetning av boligbyggingsmateriell, boliginnredning og husholdningsutstyr. Boligbyggingen må få sin egen sentralorganisasjon som ved å dirigere lån og bidragsytelse og ved faglig rettleiing og kontroll skal søke å virkeliggjøre dette målet: Et ordentlig hjem for hver husholdning i landet. Alle arbeidende nordmenn, selv de med de laveste inntekter, må få en fullverdig bolig til en pris som de kan betale”(DNA party program 1945, Fjerstad 2001). “Arbeiderpartiet vil ikke la markedskreftene rå grunnen alene på dette området.” (DNA party program 2005-9, Fjerstad 2001). 144Riksbyggen is still owned to a considerable extent by LO (2005: 41%; other organisations owning parts of Riksbyggen are e.g. KF 3%, Folksam 1%; the single housing cooperatives own 54%). 145 “I et markedssamfunn må man til enhver tid vaere konkurransdyktige. Det er derfor viktig at boligbyggelagene til enhver tid har tilbud som medlemmene önsker. Også innenfor forvaltningsvirksamheten er det derfor viktig at man stadig blir bedre, samtidigt som det skjer en utvikling I den retning medlemmene önsker.” “OBOS is now investing considerable sums in vacant sites and housing projects. It is leveraging its financial strength and market dominance to negotiate on behalf of its 197,000 members to bring them advantages such as cheaper electricity, insurance, telecommunications, oil, cable TV as well as good offers on cultural happenings and home products.”

255

146 Data for the early decades are not available for Norway due to a non- specification of cooperative dwellings or households in public data. 147In 1988, household income of cooperative occupants was 196 000 nrk, slightly below the average income of 213 000 nrk of all households (the income for home- owners was 243 000 and for renting tenants 142 000; NOS Survey of Housing Conditions 1988). 148 However, the difference in renting tenants’ voting increased. In other words, cooperative owners reduced their Labour Party voting to a lesser degree than renting occupants. Renting tenants were more strongly affiliated with the communist parties (for a study of the division between communists and Social Democratic Party over tenure preferences, see Gulbrandsen 1982 for Oslo and Nagel 1992 for Bergen). 149 A somewhat different support pattern of cooperative inhabitants is indicated for the largest cities. Bay shows that the support for the social democrats in the largest cities decreased to a stronger degree in 1981 (Bay 1985). Likewise, a member study of Oslo Bygge - Og Sparelag, the largest regional housing cooperation in Norway, indicates a decrease of Labour Party support. Fifty- four per cent of persons who became members in OBOS before the 1950s considered themselves Labour Party affiliates. However, this affiliation pattern changed over the decades and members in the 1970s no longer considered themselves only as Labour voters, but also as Conservative voters (Gulbrandsen 1980). 150 The extent to which these religious minorities pursued a competitive strategy or a more defensive strategy to safeguard their specific cultures remains an open question. However, due to the decline in their organisations since the 1940s, the strategy of these groups is here understood as defensive rather than competitive. 151 The Norwegian State Church is a public organisation but is independent with respect to welfare activities. 152 In Sweden, right-wing parties enacted full eligibility for, first, statutory and, then, municipal subsidies for all associations. Furthermore, they opposed the introduction of maximum levels of parent fees. 153 On one side, the leadership of cooperatives abolished price restrictions tacitly because they assumed that they would be unable to uphold them (Lundevall 1996). On the other side, they reintroduced some restrictions, showing that they still felt that they could rely on the support of inhabitants for such measures. However, these reintroduced restrictions were then abolished by political interference. 154 Party membership as a percentage of the total electorate dropped in Norway from 16 to 8% and in Sweden from 24 to 7% from the 1960s to the 1990s. Membership bases also became less relevant for candidates’ careers and socialisations, due to parties’ increased reliance on candidates with no long-standing party membership (Svåsand 1994; Pierre and Widfeldt 1994). With diminished memberships, parties cannot rely to the same degree on members’ dues or voluntary help for communicating party policies. However, advantages of membership bases are less relevant for party elites, due to the increase in statutory finances, the role of modern media in addressing voters, and the deliberate choice to recruit candidates from outside parties (Mair 1997). 155The basic concept Rokkan uses is cleavage structure. Cleavages are major social conflicts, but are distinguished from the multitude of conflicts by their impact

256

on a society. Cleavages are conflicts that have mobilised individuals and led to the formation of organisations. Rokkan’s concept of cleavages thus recognizes the importance of socio-political organisation of conflicting interests. (Rokkan concentrated mainly on political parties as organisations founded along the conflict lines of societies). The term cleavage structure accounts for the hierarchy of single cleavages. Cleavage structures are patterns of cleavages with more or less mobilised, organised cleavages. Even if several conflicts are socio-politically established, they are to different degrees socially influential. Social groups are to varying degrees able to establish their interests in ways that give them priority or dominance in relation to other groups’ interests. When cleavages weaken organisations, due to overlapping interests and fragmentation, Rokkan calls this the “cross-cutting” of cleavages (cp. Rokkan et al. 2001). (Cleavages in historical succession tend to “cross-cut” each other. Older cleavages eventually narrow the mobilisation market of newer conflicts. The difficulty for newer conflict mobilisation is that these conflicts are confronted with already existing individual priorities, organisations and institutions. In cases where both older and newer conflicts have persistent social support, conflicts “cross-cut” each other. That means that not all advocates of a newer conflict support one specified organisation, but several organisations, each representing some form of compromise between priorities of older and newer conflicts. For example, the conflict between work and capital gained no complete priority vis-à-vis older • conflicts. Denominational parties and left-liberal parties are representatives of compromises concerning the conflicts around nation-building and industrialisation). 156 Originally, laymen and dissidents were a countermovement opposing the State Church and its related libertarian religious orientation. This conflict situation changed after the Second World War and dissidents and State Church Protestants settled their conflict. This can be interpreted as a strategic choice when confronted with a predominantly secular society (Botvar and Aagedal 2002). 157 School education developed from predominantly religious education to an education in subjects such as mathematics, natural sciences and languages. This was the overall trend in all European countries after British industrialisation. Educational contexts changed and the task of regulation was gradually transferred to statutory bureaucracies. European churches lost their primacy in education. The resulting conflicts and compromises vary between countries. Norway differs from Sweden in having a compulsory Christian orientation, private religious schools and Christian religion taught as a school subject. In summary, the State Church and religious groups could preserve a stronger influence on school education in Norway than in Sweden. The opponents in both countries were religious groups and their related parties, on the one side, and the worker movement with related parties, on the other. The Labour Party in Norway as in Sweden tried to secularise school education in content and organisation, but was less successful in implementing their aim. Crucial for the relationship between the Labour Party and the State Church was Labour’s attempt to secularise school teaching about religion since the 1920s (Bondevik 1975, Bull 1979). 158 Rokkan regards different timing of democracy and industrialisation as a reason for the different levels of success of Swedish and Norwegian dissident

257 movements. In Sweden, the labour movement and the dissident movement were not represented in parliament when industrialisation occurred, while in Norway the dissident movement was already represented by Venstre. In Sweden, the labour movement mobilised together with dissident and other counter-cultural movements for enfranchisement, while in Norway the labour movement had to mobilise out of opposition (Rokkan and Hagvet 1987; also Lundkvist 1977). Another interesting attempt explains the difference between Norway and Sweden due to geographical concentration. The peripheral counter-culture was more strongly organised in Norway due to a greater geographical concentration. Mobilisation of the counter-culture accumulated in one geographic area in Norway, the Southwest. This may have been crucial for the development of a political organisation, gathering all these cultural interests apart from other organisations. In Sweden, the different interests in free churches and temperance had a dispersed regional pattern: the temperance movement was highly mobilised in the north, but the free churches dominated in the so-called “Bible belt” across the southern part of the country (Lindström 1979). A third attempt deals with the timing of the introduction of democracy and organisational inertia, which varies slightly with Rokkan’s. General suffrage was introduced in Norway before industrialisation. The related mobilisation for elections was traditional-conservatively organised without concurrence through working-class mobilisation. Traditional-conservative organisations may afterwards have had an advantage, due to this formation phase of political mobilisation. Early democratic organisation, based on pre-industrial interests, may have influenced the organisation of later interests that arose around the work-capital conflict. Another suggestion sees the experience of the Second World War as a reason for a stronger religiosity. Today, differences in religiosity seem not to be large: in both countries around 3% of the population are regular churchgoers, and opinion surveys suggest that around 10% of the population are convinced believers (Winsnes 2002, Dahlgren and Gustafsson 1985). On the other side, religious organisations may have gained respectability through their active resistance during the occupation years (Bull 1979), which encouraged passive acceptance of their activities. 159Historically, the special cultural orientation in the Southwest is not connected solely to religiosity, even if religiosity is a central aspect (Øidne 1957). Several popular movements (folkrörelser) had their origins or strongholds in the Southwest, such as the temperance movement, the movement for a traditional Norwegian language (nynorsk) and the lay movement. 160 Exceptions to this trend are seen in environmental organisations, which have a clear ideological and societal orientation and work in interaction with other groups. However, they could not gain the same organisational influence as former popular movements. 161 In Sweden, SAF may have been more interested in a solution with stronger immigration of employees in the 1960ies according to Lundh and Olsson. SAF argued that women had shown to be reluctant of entering the labour market in the 40ies and that a solution through foreign employees was cheaper. Why LO preferred a home-female work force solution is not clear. Maybe the higher organisation rate in unions of Swedish female employees in comparison with foreign employees was a reason. Another reason may have been that female workers

258

were not standing in direct concurrence towards male Swedish workers due to segregated labour market participation patterns (Lundh and Ohlsson 1994). 162 The Labour Market Commission was a central organisation, where the state, employers and unions met. It was the main organisation of Swedish corporatism. 163 “Under rådande förhållanden med brist på arbetskraft är det av stor betydelse att tillgängliga arbetskraftreserver kunna tagas i anspråk. Då kvinnorna utgöra den reserv, som allt mer måste fylla luckorna inom arbetslevet, framstår det som synnerligen angeläget, att de hinder, som kunna försvåra för kvinnorna att gå i produktionslivet, i möjligast mån undanröjas. Kommissionen har i olika sammanhang erfarit, att svårigheterna för mödrarna att placera sina barn många gånger gör det omöjligt för dem att antaga arbete. […] Situationen på arbetsmarknaden påkallar uppenbarligen åtgärder för en utökning av daghems- och barnträdgårdsverksamheten, och utvecklingen kommer med all sannolikhet att ytterligare markera behovet dä rav att medföra skärpta krav på kvinnornas arbetsinsatser inom produktionslivet. Genom att stimulera och understödja ifrågavarande verksamhet skulle staten befrämja ett ianspråktagande av latent kvinnlig arbetskraft, vilket enligt arbetsmarknadskommissionens mening är ett samhällsinteresse av största betydelse.” (Statens arbetsmarknadskommissionen 1942). 164 The only organisation, which opposed a further expansion of day-care, was the Union of Academics, SACO. 165 ”att det mot bakgrunden av den arbetsmarknadssituationen, som kunde • förväntas för de närmaste åren, och den ökning av yrkesverksamheten bland mödrarna, som sannolikt kunde påräknas, vore angeläget, att frågan om omfattningen och utformningen av statens stöd till barnstugor och liknande ändamål bleve föremål för en allmän översyn. […] Därjämte borde så långt som möjligt söka bedömas det framtida behovet av daghemsplatser. Utskottet förutsatte, att denna översyn skulle kunna bedrivas skyndsamt.” (Statutskottetes utlåtande, quoted in Prop. 1963:62). 166 The development was not as dramatic as expected due to the international oil crises, which also had negative repercussions on the Norwegian economic. Nevertheless, the situation of the labour market remained tightened and uncertain due to oil production.

259 References

Aardal B, Valen H. 1997. The Storting elections of 1989 and 1993: Norwegian politics in perspective. In Challenges to political parties: the case of Norway, ed. K Strøm, L Svåsand. Ann Arbor, Mich.: University of Michigan Press.

Ahlbäck S, Rothstein B, eds. 2001. Politik som organisation: förvaltningspolitikens grundproblem. Stockholm: Studieförbund Näringsliv och samhälle.

Ahrne G. 1990. Agency and organization: towards an organizational theory of society. London: Sage.

Alber J. 1987. Vom Armenhaus zum Wohlfahrtsstaat: Analysen zur Entwicklung der Sozialversicherung in Westeuropa. Frankfurt: Campus.

Alber J. 1995. A framework for the comparitive study of social services. Journal of European Social Policy (5): 131-149.

Allardt E, Rokkan S, eds. 1970. Mass politics: studies in political sociology. New York: Free press.

Andrén N, Allardt E. 1981. Nordic democracy: ideas, issues, and institutions in politics, economy, education, social and cultural affairs of Denmark, Finland, Iceland, Norway, and Sweden. Copenhagen: Det danske selskab.

Anheier HK, Seibel W, eds. 1990. The third sector: comparative studies of nonprofit organizations. Berlin: de Gruyter.

Annaniassen E. 1991. Boligsamvirkets historie i Norge. Bd. 1, Hvor Nr. 13 ikke er. Oslo: Gyldendal.

Annaniassen E. 1996a. Boligsamvirkets historie i Norge. Bd. 2, Nå bygger vi den nye tid. Oslo: Exil forlag.

Annaniassen E. 1996. Boligsamvirkets historie i Norge. Bd. 3, Tidene skifter. Oslo: Exil Forlag.

260

Annaniassen E. 2006. Norge - det socialdemokratiska ägarlandet. In Varför så olika?: nordisk bostadspolitik i jämförande historisk ljus, ed. B Bengtsson, E Annaniassen. Malmö: Égalité.

Antman P. 1996. Barn och äldreomsorg i Tyskland och Sverige. Stockholm: Välfärdsprojektet Socialdep.

Archambault E. 1997. The nonprofit sector in France. Manchester: Manchester University Press.

Backhaus-Maul H, Olk T, Bank für Sozialwirtschaft. 1995. Von Subsidiarität zu "outcontracting:“ zum Wandel der Beziehungen zwischen Staat und Wohlfahrtsverbänden in der Sozialpolitik. Berlin: Bank für Sozialwirtschaft.

Badelt C. 1989. Government versus Private Provision of Social Services: The Case of Austria. In The Nonprofit sector in international perspective: studies in comparative culture and policy, ed. E James. New York: Oxford University Press.

• Balchin P. 1996a. Introduction. In Housing policy in Europe, ed. P Balchin. London: Routledge.

Balchin P, ed. 1996b. Housing policy in Europe. London: Routledge.

Baldwin P. 1990. The politics of social solidarity: class bases of the European welfare state 1875-1975. Cambridge: Cambridge Univ. press.

Ball M, Harloe M, Martens M. 1988. Housing and social change in Europe and the USA. London: Routledge.

Barbetta GP, ed. 1997. The nonprofit sector in Italy. Manchester: Manchester University Press.

Barlow J, Duncan SS. 1994. Success and failure in housing provision: European systems compared. Oxford: Pergamon.

Bartolini S, Mair P. 1990. Identity, competition and electoral availability: the stabilisation of European electorates 1885-1985. Cambridge: Cambridge University Press.

261 Bay A-H. 1985. Boligstatus og stemmegivning: eller boligpolitikkens betydning for høyrebølgen. Oslo: Universitetet.

Bengtsson B. 1992. Not the Middle Way But Both Ways - Cooperative Housing in Sweden. In Policy, organization, tenure: a comparative history of housing in small welfare states, ed. LJ Lundqvist. Oslo: Scandinavian University Press.

Bengtsson B. 1993. Prosperous ambivalence: housing cooperatives in Sweden. Gävle: National Swedish Institute for Building Research.

Bengtsson B. 2001. Bostaden som social rättighet. In Den nya bostadspolitiken, ed. A Lindbom. Umeå: Boréa.

Bengtsson B. 2006. Sverige - kommunal allmännytta och korporativa särinteressen. In Varför så olika?: nordisk bostadspolitik i jämförande historisk ljus, ed. B Bengtsson, E Annaniassen. Malmö: Égalité.

Bengtsson B, Annaniassen E, eds. 2006. Varför så olika? Nordisk bostadspolitik i jämförande historisk ljus. Malmö: Égalité.

Bergqvist C. 1999. Likestilte demokratier? Kjønn og politikk i Norden. Oslo: Universitetsforl. i samarbeid med Nordisk ministerråd.

Beveridge W, Deakin N. 2000. Voluntary action London: Routledge.

Beyme K. v. 1982. Parteien in westlichen Demokratien. München: Piper.

Blanc M. 1993. Housing segregation and the poor: new trends in French social rented housing. Housing Studies 8:217-3.

Blom-Hansen J. 1998. Fuld behovsdækning! Skandinavisk børnepasningspolitik mod år 2000. Nordisk administrativt tidsskrift 79:329-53.

Blom-Hansen J, Henneberg S. 1998. Børnepasningsdækning i de skandinaviske lande en ajourført opgørelse. Aarhus: Department of Political Science.

Blumenroth U. 1975. Deutsche Wohnungspolitik seit der Reichsgründung: Zusatz Darstellung und kritische Würdigung Münster: Selbstverl. des Inst. für Siedlungs- und Wohnungswesen.

262 Bo (editorial staff). 2005. Ny kurs i boligpolitiken? Bo 2005.

Boelhouwer P, van der Heijden H. 1992. Housing systems in Europe. Part 1, Comparative study of housing policy. Delft: Delft Univ. Press.

Bondevik K. 1975. Kristent innslag i norsk politikk 1800-1930. Oslo: Luther.

Bostadsstyrelsen. 1968. Housing in the Nordic countries. Copenhagen: Dansk bibliografisk kontor.

Botvar PK, Aagedal O. 2002. Ost er ost og vest er vest...? In Tallenes tale 2002: perspektiver på statistikk og kirke, ed. OG Winsnes. Trondheim: Tapir.

Boverket. 1993. Svensk bostadsmarknad i internationell belysning. Boverket, Karlskrona.

Boverket. 1994. Bostadsmarknaden och 90-talets förändringar. Boverket, Karlskrona.

• Boverket. 1995. Bostadsmarknaden, Danellsystemet, krisårgångarna. Boverket, Karlskrona.

Boverket. 2005. Många mål - få medel: Boverkets utvärdering av statliga stöd till bostadsbyggandet 1993-2004: en sammanfattning av rapporten, Boverket, Karlskrona.

Bull E. 1979. Norges historie. Oslo: Cappelen.

Curtis J, Baer D, Grabb E. 2001. Nations of joiners: Explaining voluntary association membership in democratic societies. American Sociological Review 66:783-805.

Curtis J, E. G, Baer D. 1992. Voluntary association membership in 15 coutnries - a comparative analysis. American Sociological Review 57:139-52.

Dahlgren C, Gustafsson G. 1985. Religiös förändring i Norden 1930-1980. Malmö: LiberFörlag.

263 Danermark B, Elander I. 1994a. Introduction. In Social rented housing in Europe: policy, tenure and design, ed. B Danermark, I Elander. Delft: Delft Univ. Press.

Danermark B, Elander I, eds. 1994b. Social rented housing in Europe: policy, tenure and design. Delft: Delft Univ. Press.

Defourny J, Monzon Campos JL, eds. 1992. Économie sociale: entre économie capitaliste et économie publique. Bruxelles: De Boeck- Wesmael.

DNA. http://www.dna.no. Oslo: Den Norske Arbeiderpartiet (DNA).

Donnison D, Ungerson C. 1982. Housing policy. Harmondsworth: Penguin.

Ds S. 1965:3. Familjeberedningens PM angående utbyggnad av samhällets barntillsyn. Stockholm: Socialdep.

Ds U. 1985:5. Förskola och skola: en historisk återblick: en rapport från Förskola-skola-kommittén. Stockholm: Liber/Allmänna förl.

Duverger M. 1967. Political parties - their organization and activity in the modern state. London: Methuen.

Ebbinghaus B, Visser J. 1999. When institutions matter - Union growth and decline in Western Europe, 1950-1995. European Sociological Review 15:135-58.

Ebbinghaus B, Visser J. 2000. Trade unions in Western Europe since 1945. Basingstoke: Macmillan.

Ebbingshaus B. 1993. Labour unity in union diversity. Trade unions and social cleavages in Western Europe, 1890-1989. Florence: EUI.

Elander I. 1994. Paradise lost? Desubsidization and social housing in Sweden. In Social rented housing in Europe: policy, tenure and design, ed. B Danermark, I Elander. Delft: Delft Univ. Press.

Engberg J. 1986. Folkrörelserna i välfärdssamhället. Umeå: Umeå Universitet.

Esping-Andersen G. 1985. Politics against markets: the social democratic road to power. Princeton, N.J.: Princeton Univ. Press.

264

Esping-Andersen G. 1990. The three worlds of welfare capitalism. Cambridge: Polity.

Esping-Andersen G. 1999. Social foundations of postindustrial economies. New York: Oxford University Press.

Evans G. 2004. Voting: Class. In International Encyclopedia of the Social & Behavioral Sciences, ed. NJ Smelser, PB Baltes, pp. 16311-4. Amsterdam: Elsevier.

Evans PB, Rueschemeyer D, Skocpol T. 1985. Bringing the state back in. Cambridge: Cambridge University Press.

Evers A, Laville J-L, eds. 2004. The Third sector in Europe. Cheltenham: Edward Elgar.

Finlayson G. 1994. Citizen, state, and social welfare in Britain, 1830-1990. Oxford: Clarendon Press.

Fischer T. 1995. Die europäische Genossenschaft. Münster: Regensberg. •

Fjerstad LI. 2001. Vi vil-! Norske partiprogrammer 1884-2001 elektronisk ressurs. Bergen: Norsk samfunnsvitenskapelig datatjeneste, Elektronisk arkiv av norske partiprogrammer. 1 optisk plate (CD- ROM).

Flora P, ed. 1986. Growth to limits the Western European welfare states since World War II. Berlin New York: de Gruyter.

Flora P, Heidenheimer AJ, eds. 1995. The development of welfare states in Europe and America. New Brunswick, N.J.: Transaction.

Flora P, Kraus F, Pfenning W. 1983. State, economy and society in Western Europe 1815-1975 a data handbook in two volumes. Frankfurt am Main: Campus.

Flora P, Neil JS, Paul BB. 2001. Rokkan, Stein (1921-79). In International Encyclopedia of the Social & Behavioral Sciences, ed. NJ Smelser, PB Baltes, pp. 13390-4. Oxford: Pergamon.

Flora P, Alber J. 1995. Modernization, Democratization, and the Development of Welfare States in Western Europe. In The

265 development of welfare states in Europe and America, ed. P Flora, AJ Heidenheimer. New Brunswick, N.J.: Transaction.

Folin M. 1987. Housing Policies in West European Countries: The Beginning. In Between state and market: housing in the post- industrial era, ed. Lundqvist LJ, Kemeny J, Turner B. Stockholm: Almqvist & Wiksell International.

Forbrukerombudet. http://www.forbrukerombudet.no.

Freedman DA. 2004. Ecological Inference. In International Encyclopedia of the Social & Behavioral Sciences, ed. NJ Smelser, PB Baltes, pp. 4027-30. Amsterdam: Elsevier.

Föreningen Kooperativa Studier. http://www.koopi.se/fks. Stockholm: Föreningen Kooperativa Studier.

Föreningen Kooperativa Studier. various years. Kooperativ årsbok. Stockholm: Föreningen Kooperativa studier

Gaskin K, Smith JD. 1995. A new civic Europe? A study of the extent and role of volunteering: Volunteer Centre UK.

Gesamtverband Gemeinnütziger Wohnungsunternehmen. various years. Gesamtverband Gemeinnütziger Wohnungsunternehmen Wohnungswirtschaftliches Jahrbuch. Hamburg: Hammonia-Verlag.

Gidron B, Kramer RM, Salamon LM, eds. 1992. Government and the third sector: emerging relationships in welfare states. San Francisco: Jossey-Bass.

Gillström Å, ed. 1995. Den samtida historien: studier till Gävles 550- årsjubileum 1996. Stockholm: Carlsson.

Gjems-Onstad O. 1995. Rettslige rammebetingelser for tredje sector i Norge og Norden. In Frivillig organisering i Norden, ed. P Selle, KK Klausen. Oslo: TANO.

Good HG, Teller JD. 1969. A history of Western education. New York: Macmillan.

Goodin RE, Klingemann H-D, Oxford University Press., eds. 1998. A New Handbook of Political Science. Oxford: Oxford University Press.

266

Green A. 1990. Education and state formation: the rise of education systems in England, France, and the USA. New York: St. Martin's Press.

Grindheim JE, Selle P. 1989. The role of voluntary social welfare organizations in Norway: a democratic alternative to a bureaucratic welfare state? Bergen: LOS-senter.

Gulbrandsen L. 1982. Fra marked til administrasjon? Boligmarked og boligpolitikk i Oslo i det tjuende århundre. Oslo: Institutt for anvendt sosialvitenskapelig forskning.

Gulbrandsen L. 1983. Popularitetens fordelningspolitiske pris: Om endringer i den sosiale sammensetningen av medlemsmassen i OBOS i etterkrigstida. Tidsskrift for arbeiderbevegelsens historie 1983:79-96.

Gulbrandsen L. 1988. Norway. In Between owner-occupation and rented sector: housing in ten European countries, ed. Hans Kroes, Frits Ymkers, A Mulder. Delft: De Bilt: The Netherlands Christian Institute for Social Housing NCIV.

• Gustafson L. 1974. HSB under femtio år: en organisationsstudie. Stockholm: HSB:s riksförb.

Gustafsson G. 1985. Utvecklingslinjer på det religiösa området i de nordiska länderna: en jämförelse. In Religiös förändring i Norden 1930-1980, eds. Curt Dahlgren, Gustafsson G. Malmö: LiberFörlag.

Gustafsson G. 2000. Tro, samfund och samhälle: sociologiska perspektiv. Örebro: Libris.

Haakfort J, van Leuvensteijn M, Renes G. 2002. Woningcorporaties: prikkels voor effectiviteit en efficiëntie. GM 's Gravenhage: Centraal Planbureau.

Hansmann H. 1987. Economic Theories of Nonprofit Organizations. In The nonprofit sector: a research handbook, ed. WW Powell. New Haven: Yale Univ. Press.

Harloe M. 1995. The people's home? Social rented housing in Europe & America. Oxford: Blackwell.

267 Hatje A-K. 1995. Folkbarnträdgården i Norden - det goda hemmets politik. In På tröskeln till välfärden, ed. M Taussi-Sjöberg, T Vammen.

Hatje A-K. 1999. Från treklang till triangeldrama: barnträdgården som ett kvinnligt samhällsprojekt under 1880-1940-talen. Lund: Historiska media. Hawkesworth M, Kogan M, eds. 1992. Encyclopedia of government and politics. Vol. 1. London: Routledge.

Headey B. 1978. Housing policy in the developed economy: the United Kingdom, Sweden and the United States. London: Croom Helm.

Helgesson V. 1995. Debatten om dagis: barnomsorg i Gävle 1940-1975. In Den samtida historien: studier till Gävles 550-årsjubileum 1996 ed. Å Gillström. Stockholm: Carlsson.

Hills J. 1989. The Voluntary Sector in Housing: The Role of Brittish Housing Associaitons. In The Nonprofit sector in international perspective: studies in comparative culture and policy, ed. E James. New York: Oxford University Press.

Hinnfors J. 1992. Familjepolitik: samhällsförändringar och partistrategier 1960-1990. Stockholm: Almqvist & Wiksell International.

Hollingsworth J, Hollingsworth EJ. 1989. Public Organizations: The Behaviour of Hospitals in England and Wales. In The Nonprofit sector in international perspective: studies in comparative culture and policy, ed. E James. New York: Oxford University Press.

Holmans AE. 1987. Housing policy in Britain: a history. London: Croom Helm.

Holmberg S, Oscarsson H. 2004. Väljartrender: en sammanfattning av några resultat från valundersökningarna 1956-2002. Stockholm: SCB.

Holmdahl B. 2000. Tusen år i det svenska barnets historia. Lund: Studentlitteratur.

Housewives' Federation. various years. Annual reports - Årsmeldinger, Norges Kvinne- og Familieforbund, Oslo.

268 HSB. various years. Annual reports - HSB Riksförbund Årsredovisningar. Stockholm: HSB Riksförbund.

HSB. http://www.hsb.se. Stockholm: Hyresgästernas Sparkasse och Byggnadsförening (HSB).

Huber E, Stephens JD. 2001. Development and crisis of the welfare state: parties and policies in global markets. Chicago: The University of Chicago Press.

Hårsman B, Quigley JM, Anas A. 1991. Housing markets and housing institutions: an international comparison. Boston: Kluwer Academic Publ.

Häußermann H. 1994. Social housing in Germany. In Social rented housing in Europe: policy, tenure and design, ed. B Danermark, I Elander. Delft: Delft Univ. Press.

Høeg IM, Hegstad H, Winsnes OG, Kirkeforskning Stiftelsen. 2000. Folkekirke 2000: en spørreundersøkelse blant medlemmer av Den norske kirke. Oslo: KIFO - Stiftelsen Kirkeforskning. •

ILO. 1924. European Housing Problems since the War 1914-1923, International Labour Office, Genève.

ILO. 1930. Housing policy in Europe: Cheap home building, International Labour Office, Genève.

ILO. 1964. Housing co-operatives, International Labour Office, Genève.

IMF. http://www.imf.no. Straume: Indremisjonsforbundet.

James E. 1987. The nonprofit sector in comparative perspective. In The nonprofit sector: a research handbook, ed. WW Powell. New Haven: Yale Univ. Press.

James E. 1989a. Introduction. In The Nonprofit sector in international perspective: studies in comparative culture and policy, ed. E James. New York: Oxford University Press.

James E, ed. 1989b. The Nonprofit sector in international perspective: studies in comparative culture and policy. New York: Oxford University Press.

269

Jeppsson-Grassman E, Svedberg L. 1995. Frivilligt socialt arbete i Sverige - både mer och mindre. In Medmänsklighet att hyra? Åtta forskare om ideell verksamhet, ed. IE Amnå. Örebro: Libris.

Johansson H. 1980. Folkrörelserna i Sverige. Stockholm: Sober.

Judge K, Smith J. 1989. Purchase of Service in England. In The Nonprofit sector in international perspective: studies in comparative culture and policy, ed. E James. New York: Oxford University Press.

Karvonen L, Kuhnle S. 2000. Party systems and voter alignments revisited. London: Routledge.

Karvonen L, Selle P. 1995. Women in Nordic politics. Aldershot: Dartmouth.

Katz RS, Mair P. 1994. How parties organize: change and adaptation in party organizations in western democracies. London: Sage.

Kemeny J. 1995. From public housing to the social market: rental policy strategies in comparative perspective. London: Routledge.

Kendall J, Knapp M. 1996. The voluntary sector in the United Kingdom. Manchester: Manchester Univ. Press.

Kersbergen Kv. 1995. Social capitalism: a study of Christian democracy and the welfare state. London: Routledge.

Kirchheimer O. 1966. The Transformation of the Western European Party Systems. In Political parties and political development, ed. J LaPalombara, M Weiner. Princeton, N.J.: Princeton University Press.

Klausen KK, Selle P, eds. 1995b. Frivillig organisering i Norden. Oslo, København: TANO , Jurist- og Økonomforbundets Forl.

Klausen KK, Selle P. 1995a. Frivillig organisering i Norden. In Frivillig organisering i Norden, ed. P Selle, KK Klausen. Oslo: TANO.

Knapp M. 1989. Intersectoral Differences in Cost Effectiveness: Residential Child Care in England and Wales. In The Nonprofit sector in international perspective: studies in comparative culture and policy, ed. E James. New York: Oxford University Press.

270 Knutsen O. 1990. Organisering av barneomsorgen og skolebarnomsorgen i de nordiske landene: en komparativ studie av institusjoner, utdanninger og personell. Oslo: Institutt for sosialforskning.

Knutsen O. 2001. Social class, sector employment, and gender as party cleavages in the Scandinavian countries: a comparative longitudinal study, 1970-95. Scandinavian Political Studies 24:311-50.

Korpi W. 1978. Arbetarklassen i välfärdskapitalismen: arbete, fackförening och politik i Sverige. Stockholm: Prisma i samarbete med Inst. för social forskning.

Korpi W. 1981. Den demokratiska klasskampen: svensk politik i jämförande perspektiv. Stockholm: Tiden.

Korsvold T. 1997. Profesjonalisert barndom: statlige intensjoner og kvinnelig praksis på barnehagens arena 1945-1990. Trondheim: Historisk institutt, Det historisk-filosofiske fakultet, NTNU.

Korsvold T. 1998. For alle barn! Barnehagens framvekst i velferdsstaten. Oslo: Abstrakt forlag. •

Kramer RM. 1989. The Use of Government Funds by Voluntary Social Service Agencies in Four Welfare States. In The Nonprofit sector in international perspective: studies in comparative culture and policy, ed. E James. New York: Oxford University Press.

Kroes H, Frits Ymkers, Mulder A, eds. 1988. Between owner-occupation and rented sector: Housing in ten European countries. De Bilt: The Netherlands Christian Institute for Social Housing NCIV.

Kuhnle S, Selle P. 1992. Government and voluntary organizations: a relational perspective. In Needs and Welfare, ed. A Ware, RE Goodin. Aldershot: Avebury.

Kyle G. 1979. Gästarbeterska i manssamhället: studier om industriarbetande kvinnors villkor i Sverige. Stockholm: LiberFörlag.

Lane J-E, McKay D, Newton K. 1997. Political data handbook: OECD countries. Oxford: Oxford University Press.

271 LaPalombara J, Anderson J. 1992. Political parties. In Encyclopedia of government and politics. Vol. 1, ed. M Hawkesworth, M Kogan. London: Routledge.

LaPalombara J, Weiner M, eds. 1966. Political parties and political development. Princeton, N.J.: Princeton University Press.

Lawrence R. 1996. Switzerland. In Housing Policy in Europe, ed. P Balchin. London: Routledge

Lea JA. 1982. Framveksten av norsk barnehagepolitikk: en analyse av interessenter på barnehagemarkedet. Oslo: UiO/UStat.

Leira A. 1992. Welfare states and working mothers: the Scandinavian experience. Cambridge: Cambridge University Press.

Leira A. 2002. Working parents and the welfare state: family change and policy reform in Scandinavia. Cambridge: Cambridge University Press.

Lien T. 1989. Daginstitusjoner for barn i Bergen 1919-1949. Bergen: Universitet i Bergen.

Lindblad I. 1984. Politik i Norden: en jämförande översikt. Stockholm: LiberFörlag.

Lindblad I, Wahlbäck K, Wiklund C. 1972. Politik i Norden: en jämförande översikt. Stockholm: Aldus/Bonnier.

Lindbom A. 2001a. Bostadspolitiskt systemskifte: varför har nedskärningar drabbat bostadssektorn mer än andra områden? In Den nya bostadspolitiken, ed. A Lindbom. Umeå: Boréa.

Lindbom A, ed. 2001b. Den nya bostadspolitiken. Umeå: Boréa.

Lindqvist R. 1990. Från folkrörelse till välfärdsbyråkrati: det svenska sjukförsäkringssystemets utveckling 1900-1990. Lund: Arkiv.

Lindqvist R. 1998. Organisation och välfärdsstat. Lund: Studentlitteratur.

Lindström U. 1979. Helgeandsholmen and beyond: center and preriphery in Sweden. Scandinavian Political Studies 2:1-17.

272 Lipset SM, Rokkan S, eds. 1967. Party systems and voter alignments: cross- national perspectives. New York: Free Press.

Listhaug O. 1997. The Decline of Class Voting. In Challenges to political parties: the case of Norway, ed. K Strøm, L Svåsand. Ann Arbor, Mich.: University of Michigan Press.

Lorentzen H. 1994. Frivillighetens integrasjon: staten og de frivillige velferdsprodusentene. Oslo: Universitetsforl.

Lovdata. http://www.lovdata.no. Oslo: Stiftelsen Lovdata.

Lujanen M, Nordiska ministerrådet. 2004. Housing and housing policy in the Nordic countries. Copenhagen: Nordic Council of Ministers.

Lundevall O, HSB riksförbund. 1992. HSB och bostadspolitiken. 1920-talet. Stockholm: HSB:s riksförb.

Lundevall O, HSB riksförbund. 1993. HSB och bostadspolitiken. 1930-talet. Stockholm: HSB:s riksförb.

• Lundevall O, HSB riksförbund. 1994. HSB och bostadspolitiken. 1940-talet. Stockholm: HSB:s riksförb.

Lundevall O, HSB riksförbund. 1995. HSB och bostadspolitiken. 1950-1965. Stockholm: HSB:s riksförb.

Lundevall O, HSB riksförbund. 1996. HSB och bostadspolitiken. 1966-1980. Stockholm: HSB:s riksförb.

Lundevall O, HSB riksförbund. 1997. HSB och bostadspolitiken. 1981-1996. Stockholm: HSB:s riksförb.

Lundh C, Ohlsson R. 1994. Från arbetskraftsimport till flyktinginvandring. Stockholm: Studieförbund Näringsliv och samhälle.

Lundkvist S. 1977. Folkrörelserna i det svenska samhället 1850-1920. Uppsala, Stockholm: Univ., Almqvist & Wiksell international.

Lundqvist LJ. 1988. Housing policy and tenures in Sweden: the quest for neutrality. Aldershot: Avebury.

273 Lundqvist LJ. 1992a. Dislodging the welfare state?: housing and privatization in four European nations. Delft: Delft Univ. Press.

Lundqvist LJ, ed. 1992b. Policy, organization, tenure: a comparative history of housing in small welfare states. Oslo, Stockholm: Scandinavian University Press.

Lundqvist LJ. 1992c. Tenure and Property: The power dimension of welfare state involvement in nordic housing. In Policy, organization, tenure: a comparative history of housing in small welfare states, ed. LJ Lundqvist. Oslo, Stockholm: Scandinavian University Press.

Lundqvist LJ. 2001. Privatisering - varför och varför inte? In Politik som organisation: förvaltningspolitikens grundproblem, ed. S Ahlbäck, B Rothstein. Stockholm: SNS förl.

Lundqvist LJ, Kemeny J, Turner B, eds. 1987. Between state and market: housing in the post-industrial era. Stockholm: Almqvist & Wiksell International.

Lundström T, Wijkström F. 1997. The nonprofit sector in Sweden. Manchester: Manchester Univ. Press.

Mair P. 1997. Party system change: approaches and interpretations. Oxford: Clarendon.

Mair P, Oxford University Press. 1998. Party System Change. Approaches and Interpretations. Oxford: Oxford University Press.

Mair P, Plasser F, Müller WC. 2004a. Introduction. In Political parties and electoral change: party responses to electoral markets, ed. P Mair, F Plasser, WC Müller. London: SAGE.

Mair P, Plasser F, Müller WC. 2004b. Conclusion: Electoral challenges and party responses. In Political parties and electoral change: party responses to electoral markets, ed. P Mair, F Plasser, WC Müller. London: SAGE.

Mair P, Plasser F, Müller WC, eds. 2004c. Political parties and electoral change: party responses to electoral markets. London: SAGE.

March JG. 1965. Handbook of organizations. Chicago: Rand McNally.

274 Marcuse P. 1994. Property rights, tenure and ownership: towards clarity of concept. In Social rented housing in Europe: policy, tenure and design, ed. B Danermark, I Elander. Delft: Delft Univ. Press.

McCrone G, Stephens M. 1995. Housing policy in Britain and Europe. London: UCL Press.

Melby K. 1995. Kvinnelighetens strategier: Norges husmorforbund 1915- 1940 og Norges lærerinneforbund 1912-1940. Trondheim: Universitetet i Trondheim.

Merton RK. 1966. Social theory and Social Structure. New York: Free Press.

Mesmin G. 1992. Urbanisme et logement: analyse d'une crise. Paris: Presses universitaires de France.

Michels R. 1948. Political parties: A sociological study of the oligarchical tendencies of modern democracy. New York.

Miljeteig-Olssen P, Tømmerbakke ER. 1987. Fra asyl til barnehage: • barnehager i Norge i 150 år. Oslo: Universitetsforlaget.

Misgeld K, Molin K, Åmark K, Teeland J, eds. 1992. Creating social democracy: a century of the Social Democratic Labor Party in Sweden. University Park: Pennsylvania State Univ. Press.

Mishra R. 1977. Society and social policy: theoretical perspectives on welfare. London: Macmillan.

Mitchell BR. 1988. British historical statistics. Cambridge: Cambridge University Press.

Mitchell BR. 1992. International historical statistics: Europe, 1750-1988. Basingstoke: Macmillan.

Mitchell BR. 2003. International historical statistics: Europe, 1750-2000. New York: Palgrave Macmillan.

Nagel A-H. 1992. Communalism or Cooperativism? The Postwar Organization of Housing Provision in Bergen, Norway. In Policy, organization, tenure: a comparative history of housing in small

275 welfare states, ed. LJ Lundqvist. Oslo, Stockholm: Scandinavian University Press.

NBBL. http://www.boligsamvirket.no. Oslo: Norske boligbyggelags landsforbund.

NBBL. various years. Annual reports - Statistiske opplysninger om NBBLs medlemslag. Oslo: Norske boligbyggelags landsforbund.

NBBL. 1956. Statistiske undersøkelse vedrørende familier bosatt i husbankfinansierte boligbyggelagsboliger våren 1956. Oslo: Norske boligbyggelags landsforbund.

NBBL. 2006. Hvem bor i borettslag? NBBL-undersøkelse av data fra Folke- og Boligtellingen 2001. Oslo: Norske boligbyggelags landsforbund.

NBO. 1990. Bostäder och boende under 40 år: rapport till NBO- konferensen i Odense 1990, Bostadspolitisk utveckling i Norden: stödåtgärdernas effekter för samhällsekonomin och privatekonomin. Stockholm: Nordiska kooperativa och allmännyttiga bostadsföretags organisation.

Nicolaysen BB. 2001. Voluntary service provision in a strong welfare state. In Working Papers. Mannheim: Mannheimer Zentrum für europäische Sozialforschung.

Nordisk Ministerråd. 1992. Generella och selektiva medel i bostadspolitiken: användning av ekonomiska stödformer i Norden i perioden 1970- 1989. København, Stockholm: Nordisk Ministerråd, Allmänna förl.

Normisjon. http://www.normisjon.no. Oslo: Normisjon.

NOS. Population and Housing Census (Folketelling) 1960, Statistiska sentralbyrå, Oslo.

NOS. Population and Housing Census (Folke- og boligtelling) 1970, Statistisk sentralbyrå, Olso.

NOS. Population and Housing Census (Folke - og Boligtelling) 1980, Statistisk sentralbyrå, Oslo.

NOS. Population and Housing Census (Folke - og Boligtelling) 1990, Statistisk sentralbyrå, Oslo.

276

NOS. Population and Housing Census (Folke - og Boligtelling) 2001, Statistisk sentralbyrå, Oslo.

NOS. Survey of Housing Conditions (Boforholdsundersøkelsen) 1973, Statistisk sentralbyrå, Oslo.

NOS. Survey of Housing Conditions (Boforholdsundersøkelsen) 1981, Statistisk sentralbyrå, Oslo

NOS. Survey of Housing Conditions (Boforholdsundersøkelsen) 1995, Statistisk sentralbyrå, Oslo.

NOS. 1970. Byggearealstatistikk, Statistisk sentralbyrå, Oslo.

NOS. 1988. Byggearealstatistikk, Statistisk sentralbyrå, Oslo.

NOSOSCO. various years. Social security in the Nordic countries: scope, expenditure and financing. Copenhagen: NOSOSCO Nordiska socialstatistikkommittén.

• NOU. 1972:39. Førskoler. Oslo: Forvaltningstjenestene, Statens trykningskontor.

NOU. 1984:20. Framtidig organisering av fritidshjem. Oslo: Universitetsforl.

NOU. 1988:17. Frivillige organisasjoner. Oslo: Forvaltningstjenestene, Statens trykningskontor.

NOU. 1988:30. Samvirke på nye områder: fornyelse, selvhjelp og demokrati. Oslo: Forvaltningstjenestene, Statens trykningskontor.

NOU. 2002:2. Boligmarkedene og boligpolitikken: utredning fra Boligutvalget. Oslo: Statens forvaltningstjeneste Informasjonsforvaltning: Akademika distributör.

NSD. Den Norske Kirke statistikk. Bergen: NSD.

NSD. Norsk samfunnsvitenskapelig datatjeneste (Norwegian Social Science Data Services): Kommunedatabasen.

277 Nyberg A. 1995. Barnomsorgen: ett kvinnligt nollsummespel eller? In Medmänsklighet att hyra? Åtta forskare om ideell verksamhet, ed. E Amnå. Örebro: Libris.

Nygårdsmoen H. 1998. Husmødrenes barnehager 1937-1949: holdninger til barnehagene i pionerfasen. Oslo.

OBOS. http://www.obos.no. Oslo: Oslo Bolig og Sparelag (OBOS).

OECD. various years. Labour force statistics. Paris: Organisation for Economic Co-operation and Development.

OECD. various years. OECD factbook: economic, environmental and social statistics. Paris: Organisation for Economic Co-operation and Development.

OECD. 1995, 1999. OECD historical statistics. Paris: Organisation for Economic Co-operation and Development.

Olofsson G. 1990. Bostadskarriären som förmögenhetsmaskin: rapport till Expertgruppen för studier i offentlig ekonomi. Stockholm: Allmänna förl.

Olofsson G. 1995. Klass, rörelse, socialdemokrati: essäer om arbetarrörelsens sociologi. Lund: Arkiv.

Oxley MJ. 1984. Housing policy in western Europe: an economic analysis of the aims and instruments of housing policy in the United Kingdom, West Germany, France, the Netherlands, Denmark and Ireland. Wetherby: British Library.

Panebianco A. 1988. Political parties: organization and power. Cambridge: Cambridge Univ. Press.

Papa O. 1992. Housing systems in Europe. Delft: Delft Univ. Press.

Papakostas A. 1995. Arbetarklassen i organisationernas värld: en jämförande studie av fackföreningarnas sociala och historiska förutsättningar i Sverige och Grekland. Stockholm: Almqvist & Wiksell International.

278 Papakostas A. 2003. Mer organisation med färre människor och många organisationer med få frågor: en essä om politiska partier och frivilliga organisationer. SCORE rapportserie 2003.

Perrin G. 1969. Reflextions on fifty years of social security. International labour review 99:249-92.

Persson T. 2001. Sektorn som försvann! In Den nya bostadspolitiken, ed. A Lindbom. Umeå: Boréa.

Pestoff VA. 1977. Voluntary associations and Nordic party systems: a study of overlapping memberships and cross-pressures in Finland, Norway and Sweden. Stockholm: Stockholms univ.

Petersson O. 1978. Den svenske väljaren: några data från 1976 års valundersökning. Stockholm.

Pierre J, Widfeldt A. 1994. Party Organizations in Sweden: Colossuses with Feet of Clay or Flexible Pillars of Government? In How parties organize: change and adaptation in party organizations in western democracies, ed. RS Katz, P Mair. London: Sage. •

Pierson C. 1991. Beyond the welfare state? The new political economy of welfare. University Park, Pa.: Pennsylvania State Univ. Press.

Powell WW, ed. 1987. The nonprofit sector: a research handbook. New Haven: Yale Univ. Press.

Power A. 1993. Hovels to high rise: state housing in Europe since 1850. London: Routledge.

Raaum NC, ed. 1995. Kjønn og politikk. Oslo: Tano.

Rauan E. 2005. Undersøking om foreldrebetaling i barnehagar, SSB Notater 2005:23. Oslo: Statistisk sentralbyrå.

Rauch D. 2005. Institutional fragmentation and social service variations: a Scandinavian comparison. Umeå: Department of Sociology Umeå University.

Regeringen. http://www.regering.se - Sveriges regeringar under 1900-talet. Stockholm: Regeringskansliet.

279 Regini M. 1992. The future of labour movements. London: Sage.

Regjeringen.no. www.regjeringen.no - Information from the Government and Ministries. Oslo: Departementenes servicesenter.

Reiersen E, Thue E. 1996. De tusen hjem: den Norske stats husbank 1946- 96. Oslo: Ad notam Gyldendal.

Riksbyggen. various years. Annual Reports - Riksbyggens verksamhetsberättelser Riksbyggen, Stockholm.

Riksbyggen. http://www.riksbyggen.se. Stockholm: Riksbyggen.

Riksdagen. http://www.riksdagen.se. Stockholm: Riksdagen.

Riksdagen. Riksdagens protokoll jämte bihang (proceedings of the Riksdag). Stockholm: Riksdagen.

Rimlinger GV. 1971. Welfare policy and industrialization in Europe, America, and Russia. New York.

Rochester C, Hedley R, Smith JD. 1995. An introduction to the voluntary sector. London: Routledge.

Rokkan S. 1970. Citizens, elections, parties; approaches to the comparative study of the processes of development. New York,: McKay.

Rokkan S, Hagtvet B. 1987. Stat, nasjon, klasse: essays i politisk sosiologi. Oslo: Universitetsforlaget.

Rokkan S, Urwin DW, Kuhnle S, Flora P. 1999. State formation, nation- building, and mass politics in Europe: the theory of Stein Rokkan: based on his collected works. Oxford: Oxford Univ. Press.

Rokkan S, Valen H. 1964. Regional Contrasts in Norwegian Politics. In Cleavages, Ideologies and Party Systems, ed. E Allardt, Littunen. Helsinki: Westermarck Society.

Rostgaard T, Fridberg T. 1998. Caring for children and older people a comparison of European policies and practices. København: Danish National Institute of Social Research.

280 Rothstein B. 2001. Organisation som politik. In Politik som organisation: förvaltningspolitikens grundproblem, ed. S Ahlbäck, B Rothstein, Studieförbundet Näringsliv och samhälle. Stockholm: SNS förl.

Sainsbury D. 1999. Gender and welfare state regimes. Oxford: Oxford Univ. Press.

Salamon LM. 1987. Partners in Public Service: The Scope and Theory of Government-Nonprofit Relations. In The nonprofit sector: a research handbook, ed. WW Powell. New Haven: Yale Univ. Press.

Salamon LM. 1999. Global civil society: dimensions of the nonprofit sector. Baltimore, Md.: John Hopkins Center for Civil Society Studies.

Salamon LM. 2004. Global civil society. London: Eurospan.

Salamon LM, Anheier HK. 1996. The emerging nonprofit sector: an overview. Manchester: Manchester University Press.

Salamon LM, Anheier HK. 1997. Defining the nonprofit sector: a cross- national analysis. Manchester: Manchester Univ. Press. •

Salamon LM, Anheier H. 1998. Social Origins of Civil Society: Explaining the Nonprofit Sector Cross-Nationally. Voluntas 9:213-48.

Sandelin B, Södersten B. 1982. Bostadsekonomi och bostadspolitik under efterkrigstiden. In Svensk ekonomi: ett samlingsverk, ed. B Södersten. Stockholm: Rabén & Sjögren.

Sannerstedt A. 2001. Implementering - hur politiska beslut genomförs i praktiken. In Politik som organisation: förvaltningspolitikens grundproblem, ed. S Ahlbäck, B Rothstein. Stockholm: SNS förl.

SAP. http://www.socialdemokraterna.se. Stockholm: SAP.

Scott WR. 2004. Organization: Overview. In International encyclopedia of the social & behavioral sciences, ed. NJ Smelser, PB Baltes, pp. 10910-7. Amsterdam: Elsevier.

Seibel W. 1989. The Function of Mellow Weakness: Nonprofit Organizations as Problem Nonsolvers in Germany. In The Nonprofit sector in international perspective: studies in comparative culture and policy, ed. E James. New York: Oxford University Press.

281

Seip A-L. 1994. Veiene til velferdsstaten: norsk sosialpolitikk 1920-75. Oslo: Gyldendal.

Selle P. 1997. Parties and voluntary organzations: Strong or weak ties? In Challenges to political parties: the case of Norway, ed. K Strøm, L Svåsand. Ann Arbor, Mich.: University of Michigan Press Selle P, Karvonen L. 1995. Women in Nordic politics: closing the gap. Aldershot: Dartmouth.

Selle P, Kuhnle S. 1990. Frivillig organisert velferd: alternativ til offentlig? Bergen: Alma Mater.

Selle P, Øymyr B. 1995. Frivillig organisering og demokrati: det frivillige organisasjonssamfunnet endrar seg 1940-1990. Oslo: Samlaget.

SF. Stortingsforhandlinger (Proceedings of the Stortinget). Oslo: Stortinget.

SFS. Svensk författningssamling Stockholm: Fritzes.

Silenstam P. 1970. Arbetskraftsutbudets utveckling i Sverige 1870-1965. Stockholm: Industriens utredningsinstitut.

Silk L. 1948. Sweden plans for better housing. Durham, North Carolina: Duke University Press.

Sills DL. 1968. International encyclopedia of the social sciences. New York: Macmillan.

Sivesind KH. 2002. The Voluntary sector in Norway: composition, changes, and causes. Oslo: Institutt for samfunnsforskning.

Skocpol T. 1992. Protecting soldiers and mothers: political origins of social policy in the United States. Cambridge, Mass.: Belknap Press of Harvard Univ. Press.

Skolverket. various years. Barnomsorg, skola och vuxenbildning i siffror: Barn, elever och personal. Stockholm: Skolverket.

Skolverket. http://www.skolverket.se. Stockholm: Skolverket.

282 Skolverket. 2000. Tillgång och efterfrågan på barnomsorg: enkät till föräldrar med barn i åldern 1-12 år. Stockholm: Statens skolverk: Liber distribution.

Skolverket. 2003:231. Uppföljning av reformen maxtaxa, allmän förskola m.m. Stockholm: Skolverket.

Skolverket. 2005:261. Uppföljning av reformen maxtaxa, allmän förskola m.m. Stockholm: Skolverket.

Smelser NJ, Baltes PB. 2004. International encyclopedia of the social & behavioral sciences. Amsterdam: Elsevier.

Smith SR, Lipsky M. 1993. Nonprofits for hire: the welfare state in the age of contracting. Cambridge, Mass.: Harvard Univ. Press.

Socialstyrelsen. various years. Råd och anvisningar i socialvårdsfrågor, Råd och anvisningar, Socialstyrelsen anser, Allmänna råd från Socialstyrelsen. Stockholm: Kungl. Socialstyrelsen Andra byrån/ Kungl. Socialstyrelsen Socialvårdsbyrån/Socialstyrelsen.

• Solbraa-Bay J. 1970. Nordens husmødre: utgitt til Nordens husmorforbunds 50-årsjubileum. Oslo: Cappelen.

Solbraa-Bay J, Norges husmorforbund. 1985. Samfunnet sett med husmorøyne: utgitt til Norges husmorforbunds 50-års jubileum. Oslo: Norges husmorforbunds forlag.

SOS. Allmänna valen, Specialundersökningar. Stockholm, Örebro: Statistiska centralbyrån.

SOS. Byggnadsverksamheten i Sverige. Stockholm: Kungl. Socialstyrelsen.

SOS. Kooperativ verksamhet (Statistiska tabeller över till Kooperativa förbundet anslutna föreningars verksamhet). Stockholm: Komerskollegium.

SOS. Statistiska meddelanden. Serie S, Socialtjänst och socialförsäkring. Stockholm: Statistiska Centralbyrån.

SOS. various years. Bostads- och byggnadsstatistisk årsbok. Stockholm: SCB.

283 SOS. 1924. Population and Housing Census (Allmänna bostadsräkningen år) 1920. Stockholm.

SOS. 1936. Population and Housing Census 1933 (Allmänna bostadsräkningen år 1933 och därmed sammanhängande undersökningar). Stockholm.

SOS. 1952. Population and Housing Census 1945 (Bostäder och hushåll: enligt allmänna bostadsräkningen 1945 och därtill anslutna undersökningar). Stockholm: Kungl. Socialstyrelsen.

SOS. 1960. Population and Housing Census 1960 (1960 års folk- och bostadsräkning). Stockholm: Statistiska centralbyrån, Bostadsstyrelsen.

SOS. 1972. Population and Housing Census (Folk- och bostadsräkningen) 1970. Stockholm: Statistiska centralbyrån.

SOS. 1980. Population and Housing Census (Folk- och bostadsräkningen) 1980. Stockholm: Stockholm. Utrednings- och statistikkontoret.

SOS. 1990. Population and Housing Census (Folk- och bostadsräkningen) 1990. Stockholm: Statistiska centralbyrån (SCB).

SOSFS. Socialstyrelsens författningssamling. Stockholm: Socialstyrelsen.

SOU. 1938:20. Betänkande angående barnkrubbor och sommarkolonier m. m. Stockholm.

SOU. 1943:9. Utredning och förslag angående statsbidrag till daghem och lekskolor m. m. Stockholm.

SOU. 1946:63. Bostadssociala utredningen: Slutbetänkande. Stockholm,.

SOU. 1951:15. Daghem och förskolor: betänkande om barnstugor och barntillsyn. Stockholm: 1946 års kommitté för den halvöppna barnavården.

SOU. 1972:27. Förskolan: betänkande av 1968 års barnstugeutredning. Stockholm: Liberförlag/Allmänna förl.

SOU. 1979:62. Kooperationen i Sverige: betänkande. Stockholm: LiberFörlag/Allmänna förl.

284 SOU. 1981:60. Kooperationen i samhället: Kooperationsutredningens huvudbetänkande. Stockholm: LiberFörlag/Allmänna förl.

SOU. 1981:74. Från hyresrätt till bostadsrätt, LiberFörlag/Allmänna förl., Stockholm.

SOU. 1987:34. Ju mer vi är tillsammans: betänkande av 1986 års folkrörelseutredning. Stockholm: Allmänna förl.

SOU. 1992:47. Avreglerad bostadsmarknad, Utredningen om statens stöd för bostadsfinansieringen. Stockholm: Allmänna förl.

SOU. 2000:95. Utredningen om kooperativ hyresrätt. Stockholm: Fritzes offentliga publikationer.

SOU. 2001:52. Välfärdstjänster i omvandling: Antologi från Kommittén Välfärdsbokslut. Stockholm: Fritzes offentliga publikationer.

SSB. http://www.statistikkbanken.no. Oslo: SSB.

SSD. Swedish Election Studies. Gothenburg: Swedish Social Science Data • Service.

State Housing Bank. Statistics (personal communication). Oslo: Norwegian State Housing Bank.

State Housing Bank. 1997. Historisk statistikk: Den Norske Stats Husbank 1946-1996. Oslo: Norwegian State Housing Bank.

Steinmo S. 1992. Structuring politics: historical institutionalism in comparative analysis. Cambridge ; New York: Cambridge University Press.

Stjernø S. 1995. Mellom kirke og kapital: tysk velferdspolitikk - med sideblikk til britisk, svensk og norsk. Oslo: Universitetsforl.

Stortinget. http://www.stortinget.no. Olso: Stortinget.

Strandbrink P, Pestoff V. 2006. Small-scale welfare on a large scale: social cohesion and the politics of swedish childcare. Huddinge: Södertörns högskola.

285 Strøm K, Leipart JY. 1989. Ideology, Strategy, and Party Competition in Postwar Norway. European Journal of Political Research 17:263-88.

Strøm K, Svåsand L, eds. 1997. Challenges to political parties: The case of Norway. Ann Arbor, Mich.: University of Michigan Press.

Strömberg T. 1984. Kommunalsocialismen inför verkligheten: Municipal socialism and reality. Malmö: LiberFörlag.

Strömberg T. 1992. The politization of the housing market. In Creating social democracy: a century of the Social Democratic Labor Party in Sweden, ed. K Åmark, K Molin, K Misgeld, J Teeland. University Park: Pennsylvania State Univ. Press.

Strömberg T. 2001. Bostadspolitik - en historisk parentes. In Den nya bostadspolitiken ed. A Lindbom. Umeå: Boréa.

Strömberg T, Pooley CG. 1992. Housing strategies in Europe, 1800-1930. Leicester: Leicester Univ. Press. Sundell K. 1995. Barnen i förskolan: hur påverkar resurser och pedagogik barn i kommunala och privata förskolor? Stockholm: Stockholms socialtjänsts Forsknings- och utvecklingsbyrå.

Sundell K, Ståhle Y. 1996. På vinst och förlust?: hur fungerar vinstdrivna förskolor? Stockholm: Socialtjänsten: FoU-byrån.

Svenska fattigvårds-och barnavårdsförbundet. various years. Svenska fattigvårds- och barnavårdsförbundets kalender: statistisk information över den öppna barnavården. Stockholm: Norstedt.

Svensson KAS. 1993. Co-operative housing in Sweden: a building with many facades: on the diversity of co-operation. Stockholm.

Svåsand L. 1994. Change and adaptation in Norwegian party organizations. In How parties organize: change and adaptation in party organizations in western democracies, ed. RS Katz, P Mair. London: Sage.

Svåsand L, Strøm K, Rasch BE. 1997. Change and adaptation in party organization. In Challenges to political parties: the case of Norway, ed. K Strøm, L Svåsand. Ann Arbor, Mich.: University of Michigan Press.

286 Södersten B, ed. 1982. Svensk ekonomi: ett samlingsverk. Stockholm: Rabén & Sjögren.

Tomann H. 1996. Germany. In Housing policy in Europe, ed. P Balchin. London: Routledge.

Torgersen U. 1987. Housing the Wobbly Pillar under the Welfare State. In Between state and market: housing in the post-industrial era, ed.. Lundqvist LJ, Kemeny J, Turner B. Almqvist & Wiksell International.

Tranvik T, Selle P. 2007. The Rise and Fall of Popular Mass Movements Organizational Change and Globalization — the Norwegian Case. Acta Sociologica 50:57-70.

Tufte GC. 1975. Holdninger til boligenes eierforhold 1918-1974: en undersøkelse av en interessegruppes forsøk på å tilbasse seg sosial og politisk endring: Oslo: Universitetet.

Turner B. 1998. Housing cooperatives in Sweden: the effects of financial deregulation. Gävle: Institute for housing research.

• Turner B. 2001. Bostadspolitik och samhällsekonomi. In Den nya bostadspolitiken, ed. A Lindbom. Umeå: Boréa.

Turner B, Kemeny J, Lundqvist LJ, eds. 1987. Between state and market: housing in the post-industrial era. Stockholm: Almqvist & Wiksell Periodical Co.

Tuveng M. 1946. Arbeidlöshet og beskjeftigelse i Norge för og under krigen. Bergen.

Umrath H. 1953 Arbeiterbewegung und Wohnungsbau in Europa Brüssel Europ. Regional, Sekr. d. Internat. Bundes Freier Gewerkschaften.

Union HLM. http://www.union-hlm.org. Paris: L'Union sociale pour l'habitat.

United Nations. 1954. Housing through non-profit organizations. New York: United Nations, Department of Economic and Social Affairs.

United Nations. 1976. Human settlements in Europe 1976. New York: United Nations, Department of Economic and Social Affairs.

287 Urwin D. 1997. The Norwegian party system from the 1880s. In Challenges to political parties: the case of Norway, ed. K Strøm, L Svåsand. Ann Arbor, Mich.: University of Michigan Press.

Useem M. 1987. Corporate Philanthropy. In The nonprofit sector: a research handbook, ed. WW Powell. New Haven: Yale Univ. Press.

Valen H, Katz D. 1967. Political parties in Norway: a community study. Oslo: Universitetsforlaget.

Valmyndigheten. http://www.valmyndigheten.se. Solna: Valmyndigheten.

Walter M, Wächter A. 2000. Organisation och samarbete på föräldrakooperativa dag- och fritidshem. Kullavik: Föräldrakooperativens Riksförening Vårat Dagis, ADA reportage & tidningsproduktion.

Van Laarhofen P, Bart Bakker, Jaap Dronkers, Schijf H. 1990. Achievment in Public and Private Secondary Education in the Netherlands. In The third sector: comparative studies of nonprofit organizations, ed. HK Anheier, W Seibel. Berlin, New York: W. de Gruyter.

Van Til J. 1988. Mapping the third sector: voluntarism in a changing social economy. New York, NY: Foundation Center.

Ware A. 1996. Political parties and party systems. Oxford, New York: Oxford University Press.

Ware A, Goodin RE, eds. 1990. Needs and welfare. London: Sage.

Weber M. 1972. Wirtschaft und Gesellschaft. Tübingen: Mohr.

Weisbrod BA. 1977. The voluntary nonprofit sector: an economic analysis. Lexington, Mass.: Heath.

Weisbrod BA. 1990. Toward a theory of the voluntary nonprofit sector in a three-sector economy. In The Nonprofit Organization: Essential Readings, ed. DL Gies, JS Ott, JM Shafritz. Pacific Grove, California: Brooks/Cole Publishing.

Victorin A. 1987. Forms of tenure in Sweden: the case of the tenant owner. Stockholm: Almqvist & Wiksell International.

288 Victorin A. 2003. Bostadsrätt och kooperativ hyresrätt. Uppsala: Iustus.

Wiesener G. 1948. Barnevernet i Norge: en oversikt. Oslo.

Wilensky HL. 1975. The welfare state and equality: structural and ideological roots of public expenditures. Berkeley: University of California Press.

Wilensky HL, Lebeaux CN. 1958. Industrial society and social welfare: the impact of industrialization on the supply and organization of social welfare services in the United States. New York. Winsnes OG. 1998. Tallenes tale 1998: perspektiver på statistikk og kirke. Trondheim: Tapir.

Winsnes OG. 2002. Tallenes tale 2002: perspektiver på statistikk og kirke. Trondheim: Tapir.

Wollebaek D, Selle P. 2002. Det nye organisasjonssamfunnet: demokrati i omforming. Bergen: Fagbokforlaget.

Ylvisaker PN. 1987. Foundations and Nonprofit Organizations. In The • nonprofit sector: a research handbook, ed. WW Powell. New Haven: Yale Univ. Press.

Øidne G. 1957. Litt om motsetninga mellom Austlandet og Vestlandet. Oslo: Samlaget.

289

FÖRTECKNING ÖVER LICENTIATUPPSATSER OCH DOKTORS- AVHANDLINGAR VID SOCIOLOGISKA INSTITUTIONEN, UMEÅ UNIVERSITET

Licentiatuppsatser

L 1 Hass, Rita The Homophile Movement Conflict: Controversy and Change vt -68 L 2 Linné, Agneta Fruktsamhet och familjeplanering vt -68 L 3 Cottino, Amedeo Social differentiering och socialt byte: en referensram vt -70 L 4 Romàn, Ola Hastighetsöverträdelser i sociologisk belysning vt -72 L 5 Åström, Torsten Fem glesbygdsområden ur sociologisk synvinkel vt -72 L 6 Ericsson, Thomas Arbetsvärdering med poängsystem. Några lönebild- ningsteorier vt -88 L 7 Södergran, Lena Invandrar- och flyktingpolitik i praktiken - Exemplet Umeå Kommun vt -97 L 8 Lindgren, Simon Ungdomsbilder - Text och kontext i den norrbottniska mediediskursen från trettiotal till åttiotal ht -98 L 9 Miettinen, Antti Institutional Characteristics of Co-operatives Providing Welfare Services. Theoretical Study of Co-operatives and Their Members vt -00 L 10 Tigervall, Carina Tigerns avklippta morrhår. Vi och de Andra i den nya svenska filmen anno 2000 vt -03

Doctoral Theses at the Department of Sociology

D 1 Cerha, Jarko Selective Mass Communication vt -68

D 2 Boethius, Inga Individuellt och socialt betingande faktorer vid olika grupprocesser vt -70

D 3 Cottino, Amedeo Slavmarknad - eller om lagens effektivitet. En studie av den grå arbetskraften vt -73

D 4 Åberg, Rune Changes in work conditions as a result of changes in economic structure vt -73

D 5 Sahlman-Karlsson, Siiri Finska studenter i Umeå. Språkfärdighet och studie- framgång vt -74

• D 6 Hedman, Eva Boende och samhällsförändring vt -74

D 7 Pettersson, Lars-Göran Hushållens inkomst- och konsumtionsförhållanden regionala skillnader och urbaniseringseffekter vt -75

D 8 Marklund, Staffan Living Conditions and Social Policy in Rural Change vt -75

D 9 Cigéhn, Göran Stabilitet och instabilitet i partipolitiska sympatier vt -76

D 10 Wolvén, Lars-Erik Kabel-TV - Försök med ett annorlunda medium i när- samhället; sammanfattning och analys av resultaten från första försöksperioden vt -76

D 11 Eriksson, Ingalill Sociala relationer: Orsaker samt myndighetsåtgärder och sociologisk metod ht -77

D 12 Bergroth, Alf Handikappade och åldringar i en glesbygdskommun. Om effekter av socialpolitiska åtgärder vt -77

D 13 Ålund, Aleksandra Migration och sociala förändringsprocesser. Om samtida jugoslaviska arbetsmigranter vt -78

D 14 Drugge, Ulf Domstolar som konfliktreglerare. En komparativ undersökning av underrätternas konfliktreglerande verksamhet ht -78

D 15 Källtorp, Ove Transformation of social structure in peripheral village communities ht -78

D 16 Höglund, Sten Centralisering och reduktion av medlemsinflytandet i en stor facklig organisation vt -79

D 17 Höglund, Lars och Persson, Olle Kommunikation inom vetenskap och teknik vt -80

D 18 Berglund, Staffan Resisting Poverty - Perceptives on Participation and Social Development. The Case of CRIC and the Eastern Rural Region of Cauca in Colombia vt -82

D 19 Frick, Willy Strukturomvandling och social utslagning - en analys av sambanden mellan social struktur och social miss- anpassning och utslagning under perioden 1860-1975 ht -82

D 20 Lindblad, Anders Lokal radio och tv - en analys av publikstruktur och deltagande vt -83

D 21 Dahlgren, Lars Samhällsplanering och lokalsamhälle - en sociologisk analys av den sociala samhällsplaneringens möjligheter och begränsningar, illustrerad av tre ortsstudier i Norrbottens län vt -84

D 22 Höög, Jonas - Arbetstillfredsställelse och frånvaro vt -85 D 23 Lindgren, Gerd Kamrater, kollegor och kvinnor - en studie av köns- segregeringsprocessen i två mansdominerade organisationer ht -85

D 24 Gisselberg, Margareta Att stå vid spisen och föda barn. Om hushållsarbete som kvinnoarbete vt -86

D 25 Johansson, Mats Arbetararistokrater och arbetarbyråkrater - Om refor- mistiska och revolutionära tendenser inom den svenska arbetarklassen i början av seklet vt -86

D 26 Schierup, Carl-Ulrik Danser de for traditionens skyld? Invandrere, kultur og samfund vt -87

D 27 Svallfors, Stefan Vem älskar välfärdsstaten? Attityder, organiserade intressen och svensk välfärdspolitik vt -89 • D 28 Lindqvist, Rafael Från folkrörelse till välfärdsbyråkrati. Det svenska sjukförsäkringssystemets utveckling ht -90

D 29 Ekström, Marianne Kost, klass och kön ht -90

D 30 Ericsson, Thomas Systematisk arbetsvärdering: Ett lönesättnings- instrument i närbild vt -91

D 31 Halleröd, Björn Den svenska fattigdomen vt -91

D 32 Alalehto, Tage, Teknik och konflikt. LKAB 1946-1989 ht -92

D 33 Hägg, Kerstin Kvinnor och män i Kiruna. Om kön och vardag i förändring i ett modernt gruvsamhälle 1900-1990 ht -93

D 34 Schrieber, Trine Förhåbninger og skuffelser i kvindeerhvervene. Kvinders möde med ny teknologi og organisatorisk Forandring ht -94

D 35 Hansson, Carl-Gösta Fackliga karriärer. Boréa bokförlag. Umeå vt -96

D 36 Waara, Peter Ungdom i gränsland. Boréa bokförlag. Umeå vt -96

D 37 Eriksson, Nils The Psychosocial Work Environment and Illness among Office Workers vt -96

D 38 Augustsson, Gunnar Etniska relationer i arbetslivet - Teknik, arbetsorga- nisation & etnisk diskriminering i svensk bilindustri vt -96

D 39 Stern, Peter Prisoners of the Crystal Palace - Mapping & Under- standing the Social and Cognitive Organization of Scientific Research Field ht -96

D 40 Eriksson, Helén & Gunnarsson, Elly På tröskeln till omvårdnadsvärlden vt -97

D 41 Löwander, Birgitta Rasism och antirasism på dagordningen - Studier av televisionens nyhetsrapportering i början av 90-talet ht -97

D 42 Melin, Göran Co-production of Scientific Knowledge - Research Collaboration between Countries, Universities and Individuals ht -97

D 43 Stattin, Mikael Yrke, yrkesförändring och utslagning från arbetsmarknaden - en studie av relationen mellan förtidspension och arbetsmarknadsförändring vt -98

D 44 Nordenmark, Mikael Unemployment, Employment Commitment and Well-being – The Psychosocial Meaning of (Un)employment among Women and Men vt -99

D 45 Edlund, Jonas Citizens and Taxation: Sweden in Comparative Perspective vt -99

D 46 Appelqvist, Maria Responsibility in Transition. A Study of Refugee Law and Policy in Sweden vt -99

D 47 Nyman-Kurkiala, Pia Att flytta bort och hem igen. Sociala nätverk i kedjemigration vt -99

D 48 Hjerm, Mikael My People Right or Wrong? A Comparative Analysis of National Sentiments and their Meaning vt -00

• D 49 Jacobsson, Mats 'Att blifva sin egen' - Ungdomars väg in i vuxenlivet i 1700- och 1800-talens övre Norrland vt -00

D 50 Södergran, Lena Svensk Invandrar- och Integrationspolitik. En fråga om jämlikhet, demokrati och mänskliga rättigheter vt -00

D 51 Bihagen, Erik The Significance of Class. Studies of class inequalities, consumption and social circulation in contemporary Sweden ht -00

D 52 Strandh, Mattias Varying Unemployment Experiences? The economy and mental well-being ht -00

D 53 Slavnic, Zoran Existens och temporalitet. Om det samtida flyktingskapets komplexitet ht -00

D 54 Kadhim, Abdul M. Svenskt kommunalt flyktingmottagande. Politik och implementering ht -00 D 55 Bengs, Carita Looking good. A study of gendered body ideals among young people ht -00

D 56 Danell, Rickard Internationalization and Homogenization: A Bibliometric Study of International Management Research vt -01

D 57 Grape, Owe Mellan morot och piska. En fallstudie av 1992 års rehabiliteringsreform ht -01

D 58 Mellberg, Nea När det overkliga blir verklighet. Mödrars situation när deras barn utsätts för sexuella övergrepp av fäder vt -02

D 59 Nyman, Charlott Mine, yours or ours? Sharing in Swedish couples vt -02

D 60 Evertsson, Lars Välfärdspolitik och kvinnoyrken. Organisation, välfärds- stat och professionaliseringens villkor vt -02

D 61 Danielsson, Erna Är delaktighet möjlig i en byråkrati? En fallstudie inom Försvarsmakten av det arbete som föregick försvarsbeslut -96 vt -02

D 62 Nordlund, Anders Resilient Welfare States – Nordic Welfare State Develop- ment in the Late 20th century vt -02

D 63 Lindgren, Simon Modernitetens markörer. Ungdomsbilder i tid och rum vt -02

D 64 Olofsson, Anna Waves of Controversy. Gene Technology in Dagens Nyheter 1973-96 vt -02

D 65 Öhman, Susanna Public Perceptions of Gene Technology – On the Edge of Risk Society vt -02

D 66 Sundström, Eva Gender Regimes, Family Policies and Attitudes to Female Employment: A Comparison of Germany, Italy and Sweden vt -03

D 67 Mählck, Paula Mapping Gender in Academic Workplaces: Ways of Reproducing Gender Inequality within the Discourse of Equality vt -03

D 68 Karlsson, Peder Forskares socialisation: Kunskapssociologisk visit i doktoranders livsvärldar vt -04

D 69 Gustafson, Åsa Sköra livsmönster. Om integrations- och normalise- ringsprocesser bland bosniska flyktingar vt -04

D 70 Grönlund, Anne Flexibilitetens gränser. Förändring och friktion i arbetsliv och familj ht -04

• D 71 Hult, Carl The way we conform to paid Labour: Commitment to Employment and organization in a comparative perspective ht -04

D 72 Westberg, Annika Becoming an Adult. Living Conditions and Attitudes among Swedish Youth vt -05

D 73 Tigervall, Carina Folkhemsk film: med “invandraren” i rollen som den sympatiske Andre vt -05

D 74 Almqvist, Anna-Lena The Care of Children: A Cross-National Comparison of Parents´ Expectations and Experiences vt -05

D 75 Karlsson, Lena Klasstillhörighetens subjektiva dimension: Klass- identitet, sociala attityder och fritidsvanor vt -05

D 76 Rauch, Dietmar Institutional Fragmentation and Social Service Variations: A Scandinavian Comparison ht -05 D 77 Bolinder, Margareta Handlingsutrymmets betydelse för arbetslösas upp- levelser, handlingsstrategier och jobbchanser vt -06

D 78 Schmauch, Ulrika Den osynliga vardagsrasismens realitet vt -06

D 79 Jonsson, Gun Tanter och representanter. Dilemman i frivilliga organisationer: - en fråga om oligarki eller demokrati? vt -06

D 80 Kvist, Elin Stormarknadens nya maktordningar – Från kassörskor och butikschefer till (o)demokratiska arbetslag vt -06

D 81 Johansson, Gun-Britt Synderskan och Lagen: Barnamord i tre Norrlandslän 1830-1870 vt -06

D 82 Larsson, Daniel Exposure to crime as a consequence of poverty. Five investigations about relative deprivation, poverty and exposure to crime ht -06

D 83 Åström, Fredrik The Social and Intellectual Development of Library and Information Science ht -06

D 84 Bertilsdotter Rosqvist, Hanna Lagom lika, lagom olika - en diskussion om makt, retorik och bi-teoretiska/sexuella subjektiviteter vt -07

D 85 Höckertin, Chatrine Organisational characteristics and psychosocial working conditions in different forms of ownership vt -07

D 86 Grosse, Ingrid Political parties and welfare associations ht -07

Fr o m 1994 utges en ny avhandlingsserie ”Akademiska avhandlingar vid sociologiska institutionen, Umeå universitet”

A 1 Schreiber, Trine Forhåbninger og skuffelser i kvindeerhvervene. Kvinders möde med ny teknologi og organisatorisk forandring ht -94

A 2 Eriksson, Nils The Psychosocial Work. Environment and Illness Among Office Workers vt -96

A 3 Augustsson, Gunnar Etniska relationer i arbetslivet - Teknik, arbetsorga- nisation & etnisk diskriminering i svensk bilindustri vt -96

A 4 Eriksson, Helén & Gunnarsson, Elly På tröskeln till omvårdnadsvärlden vt -97

A 5 Södergran, Lena • Invandrar- och flyktingpolitik i praktiken (licentiatavhandling) vt -97

A 6 Löwander, Birgitta Rasism och antirasism på dagordningen - Studier av televisionens nyhetsrapportering i början av 1990-talet ht -97

A 7 Melin, Göran Co-production of Scientific Knowledge - Research Collaboration between Countries, Universities and Individuals ht -97

A 8 Stattin, Mikael Yrke, yrkesförändring och utslagning från arbetsmarknaden - en studie av relationen mellan förtids- pension och arbetsmarknadsförändring vt -98

A 9 Lindgren, Simon Ungdomsbilder - Text och kontext i den norrbottniska mediediskursen från trettiotal till åttiotal (licentiatavhandling) ht -98 A 10 Nordenmark, Mikael Unemployment, Employment Commitment and Well-being – The Psychosocial Meaning of (Un)employment among Women and Men vt -99

A 11 Edlund, Jonas Citizens and Taxation: Sweden in Comparative Perspective vt -99

A 12 Appelqvist, Maria Responsibility in Transition. A Study of Refugee Law and Policy in Sweden vt -99

A 13 Hjerm, Mikael My People Right or Wrong? A Comparative Analysis of National Sentiments and their Meaning vt -00

A 14 Jacobsson, Mats 'Att blifva sin egen' - Ungdomars väg in i vuxenlivet i 1700- och 1800-talens övre Norrland vt -00

A 15 Miettinen, Antti Institutional Characteristics of Co-operatives Providing Welfare Services. Theoretical Study of Co-operatives and Their Members (licentiatavhandling) vt -00

A 16 Södergran, Lena Svensk Invandrar- och Integrationspolitik. En fråga om jämlikhet, demokrati och mänskliga rättigheter vt -00

A 17 Bihagen, Erik The Significance of Class. Studies of class inequalities, consumption and social circulation in contemporary Sweden ht -00

A 18 Strandh, Mattias Varying Unemployment Experiences? The economy and mental well-being ht -00

A 19 Slavnic, Zoran Existens och temporalitet. Om det samtida flyktingskapets komplexitet ht -00

A 20 Kadhim, Abdul M. Svenskt kommunalt flyktingmottagande. Politik och implementering ht -00

A 21 Bengs, Carita Looking good. A study of gendered body ideals among young people ht -00

A 22 Danell, Rickard Internationalization and Homogenization: A Biblio- metric Study of International Management Research vt -01

A 23 Grape, Owe Mellan morot och piska. En fallstudie av 1992 års rehabiliteringsreform ht -01

A 24 Nyman, Charlott Mine, yours, or ours? Sharing in Swedish couples vt -02

A 25 Evertsson, Lars Välfärdspolitik och kvinnoyrken. Organisation, välfärds- • stat och professionaliseringens villkor vt -02

A 26 Danielsson, Erna Är delaktighet möjlig i en byråkrati? En fallstudie inom Försvarsmakten av det arbete som föregick försvarsbeslut -96 vt -02

A 27 Nordlund, Anders Resilient Welfare States – Nordic Welfare State Develop- ment in the Late 20th century vt -02

A 28 Lindgren, Simon Modernitetens markörer. Ungdomsbilder i tid och rum vt -02

A 29 Olofsson, Anna Waves of Controversy. Gene Technology in Dagens Nyheter 1973-96 vt -02

A 30 Öhman, Susanna Public Perceptions of Gene Technology – On the Edge of Risk Society vt -02 A 31 Sundström, Eva Gender Regimes, Family Policies and Attitudes to Female Employment: A Comparison of Germany, Italy and Sweden vt -03

A 32 Tigervall, Carina Tigerns avklippta morrhår. Vi och de Andra i den nya svenska filmen anno 2000 (licentiatuppsats) vt -03

A 33 Mählck, Paula Mapping Gender in Academic Workplaces: Ways of Reproducing Gender Inequality within the Discourse of Equality vt -03

A 34 Karlsson, Peder Forskares socialisation: Kunskapssociologisk visit i doktoranders livsvärldar vt -04

A 35 Gustafson, Åsa Sköra livsmönster. Om integrations- och normalise- ringsprocesser bland bosniska flyktingar vt -04

A 36 Hult, Carl The way we conform to paid Labour: Commitment to employment and organization from a comparative perspective ht -04

A 37 Westberg, Annika Becoming an Adult. Living Conditions and Attitudes among Swedish Youth vt -05

A 38 Tigervall, Carina Folkhemsk film: med “invandraren” i rollen som den sympatiske Andre vt -05

A 39 Almqvist, Anna-Lena The Care of Children: A Cross-National Comparison of Parents´ Expectations and Experiences vt -05

A 40 Karlsson, Lena Klasstillhörighetens subjektiva dimension: Klass- identitet, sociala attityder och fritidsvanor vt -05

A 41 Rauch, Dietmar Institutional Fragmentation and Social Service Variations: A Scandinavian Comparison ht -05

A 42 Bolinder, Margareta Handlingsutrymmets betydelse för arbetslösas upp- levelser, handlingsstrategier och jobbchanser vt -06

A 43 Schmauch, Ulrika Den osynliga vardagsrasismens realitet vt -06

A 44 Jonsson, Gun Tanter och representanter. Dilemman i frivilliga organisationer: - en fråga om oligarki eller demokrati? vt -06

A 45 Kvist, Elin Stormarknadens nya maktordningar – Från kassörskor och butikschefer till (o)demokratiska arbetslag vt -06

A 46 Johansson, Gun-Britt Synderskan och Lagen: • Barnamord i tre Norrlandslän 1830-1870 vt -06

A 47 Larsson, Daniel Exposure to crime as a consequence of poverty. Five investigations about relative deprivation, poverty and exposure to crime ht -06

A 48 Åström, Fredrik The Social and Intellectual Development of Library and Information Science ht -06

A 49 Höckertin, Chatrine Organisational characteristics and psychosocial working conditions in different forms of ownership vt -07

A50 Grosse, Ingrid Political parties and welfare associations ht -07